title
Watch Me Build an AI Startup

description
I'm going to build a medical imaging classification app called SmartMedScan! The potential customers for this app are medical professionals that need to scale and improve the accuracy of their diagnoses using AI. From ideation, to logo design, to integrating features like payments and AI into a single app, I'll show you my 10 step process. I hope that by seeing my thought process and getting familiar with the sequence of steps I'll demonstrate, you too will be as inspired as I am to use this technology to do something great for the world. Enjoy! Code for this video: https://github.com/llSourcell/AI_Startup_Prototype Please Subscribe! And like. And comment. That's what keeps me going. Want more education? Connect with me here: Twitter: https://twitter.com/sirajraval Facebook: https://www.facebook.com/sirajology instagram: https://www.instagram.com/sirajraval More Learning resources: https://www.youtube.com/watch?v=FTr3n7uBIuE https://www.youtube.com/watch?v=9bbS-trc8ys&t=608s https://www.youtube.com/watch?v=mrRfpiAwad0 https://www.youtube.com/watch?v=HhqhFbwiaig&t=563s Join us in the Wizards Slack channel: http://wizards.herokuapp.com/ Hit the Join button above to sign up to become a member of my channel for access to exclusive live streams! Join us at the School of AI: https://theschool.ai/ Signup for my newsletter for exciting updates in the field of AI: https://goo.gl/FZzJ5w And please support me on Patreon: https://www.patreon.com/user?u=3191693 Join my AI community: http://chatgptschool.io/ Sign up for my AI Sports betting Bot, WagerGPT! (500 spots available): https://www.wagergpt.co

detail
{'title': 'Watch Me Build an AI Startup', 'heatmap': [{'end': 694.255, 'start': 641.782, 'weight': 1}, {'end': 1615.847, 'start': 1587.677, 'weight': 0.761}], 'summary': 'Follow the journey of building an ai startup, smart medscan, including its prototype, market targeting, lean startup strategies, web app development, ai deployment, and potential future additions, aiming to reduce misdiagnoses and improve patient care in healthcare.', 'chapters': [{'end': 45.591, 'segs': [{'end': 27.017, 'src': 'embed', 'start': 0.249, 'weight': 0, 'content': [{'end': 5.051, 'text': "Hello world, it's Siraj, and who's ready to watch me build an AI startup?", 'start': 0.249, 'duration': 4.802}, {'end': 10.672, 'text': "In this video, I'm going to show you how to build an AI startup from start to finish.", 'start': 5.471, 'duration': 5.201}, {'end': 15.294, 'text': "What I mean is we're going to build a simple prototype and I have it right here for you.", 'start': 11.073, 'duration': 4.221}, {'end': 17.094, 'text': "It's called Smart MedScan.", 'start': 15.334, 'duration': 1.76}, {'end': 18.335, 'text': 'Let me show you a quick demo.', 'start': 17.114, 'duration': 1.221}, {'end': 21.336, 'text': "It's running locally on my machine as you see behind me.", 'start': 18.615, 'duration': 2.721}, {'end': 22.996, 'text': "I've called it Smart MedScan.", 'start': 21.656, 'duration': 1.34}, {'end': 27.017, 'text': "Clearly I need to design it a little better, but it's a prototype.", 'start': 23.056, 'duration': 3.961}], 'summary': 'Siraj demonstrates building an ai startup, showcasing smart medscan prototype in a video.', 'duration': 26.768, 'max_score': 0.249, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V4249.jpg'}], 'start': 0.249, 'title': 'Building an ai startup: smart medscan', 'summary': 'Details the process of building an ai startup, showcasing a prototype called smart medscan and outlining the entire pipeline of ideation, business plan design, acquiring first clients, and startup growth.', 'chapters': [{'end': 45.591, 'start': 0.249, 'title': 'Building an ai startup: smart medscan', 'summary': 'Details the process of building an ai startup, showcasing a prototype called smart medscan and outlining the entire pipeline of ideation, business plan design, acquiring first clients, and startup growth.', 'duration': 45.342, 'highlights': ['The chapter details the process of building an AI startup, showcasing a prototype called Smart MedScan and outlining the entire pipeline of ideation, business plan design, acquiring first clients, and startup growth.', 'Siraj demonstrates a prototype AI startup called Smart MedScan, emphasizing the need for design improvements and providing a quick demo running locally on his machine.', 'Siraj outlines the entire pipeline of building an AI startup, detailing the ideation, business plan design, acquisition of first clients, and startup growth over time.']}], 'duration': 45.342, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V4249.jpg', 'highlights': ['The chapter details the process of building an AI startup, showcasing a prototype called Smart MedScan and outlining the entire pipeline of ideation, business plan design, acquiring first clients, and startup growth.', 'Siraj demonstrates a prototype AI startup called Smart MedScan, emphasizing the need for design improvements and providing a quick demo running locally on his machine.', 'Siraj outlines the entire pipeline of building an AI startup, detailing the ideation, business plan design, acquisition of first clients, and startup growth over time.']}, {'end': 397.332, 'segs': [{'end': 74.215, 'src': 'embed', 'start': 45.591, 'weight': 0, 'content': [{'end': 48.596, 'text': "I'm gonna show you, give you all of my secrets of how I would do this.", 'start': 45.591, 'duration': 3.005}, {'end': 52.926, 'text': "I'm gonna be totally transparent and open with you, alright, so Check this out.", 'start': 48.596, 'duration': 4.33}, {'end': 56.088, 'text': 'So the first step for a user is to sign up for Smart MedScan.', 'start': 52.986, 'duration': 3.102}, {'end': 62.65, 'text': 'And who is this for? This is a tool for doctors to be able to classify medical images.', 'start': 56.368, 'duration': 6.282}, {'end': 65.391, 'text': 'So x-ray scans for their lungs.', 'start': 62.951, 'duration': 2.44}, {'end': 74.215, 'text': "Doctors have all sorts of images that they spend many years learning how to classify with their brain, but there's a shortage of doctors out there.", 'start': 65.912, 'duration': 8.303}], 'summary': 'Smart medscan helps doctors classify x-ray scans for lung, addressing shortage of doctors.', 'duration': 28.624, 'max_score': 45.591, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V445591.jpg'}, {'end': 137.896, 'src': 'embed', 'start': 108.537, 'weight': 2, 'content': [{'end': 112.679, 'text': "Okay, so I'm a doctor, I have an image, I wanna classify it, so I'm gonna pay with my card.", 'start': 108.537, 'duration': 4.142}, {'end': 120.343, 'text': "I'm gonna say, you know, this is my email, here's my test card number, whatever it is, it's fake, and then here's my CVC.", 'start': 112.699, 'duration': 7.644}, {'end': 123.104, 'text': "I'm gonna pay 10 bucks for a single classification.", 'start': 120.763, 'duration': 2.341}, {'end': 131.031, 'text': 'I go ahead and pay, the charge goes through, And it says, great job, thanks for paying 10 bucks.', 'start': 123.124, 'duration': 7.907}, {'end': 132.732, 'text': 'Go ahead and upload your file.', 'start': 131.071, 'duration': 1.661}, {'end': 133.573, 'text': 'This now shows up.', 'start': 132.752, 'duration': 0.821}, {'end': 137.896, 'text': "And I'm gonna say, let's see.", 'start': 134.313, 'duration': 3.583}], 'summary': 'A doctor paid 10 bucks for image classification and uploaded a file.', 'duration': 29.359, 'max_score': 108.537, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V4108537.jpg'}, {'end': 192.82, 'src': 'embed', 'start': 160.449, 'weight': 3, 'content': [{'end': 164.25, 'text': 'It is very likely that this is an example of pneumonia.', 'start': 160.449, 'duration': 3.801}, {'end': 170.591, 'text': 'This is something that people go to school for years to do, and we just figured out how to automate that, which is amazing.', 'start': 164.77, 'duration': 5.821}, {'end': 172.032, 'text': "It's gonna help doctors and help patients.", 'start': 170.611, 'duration': 1.421}, {'end': 173.332, 'text': "It's gonna make life better.", 'start': 172.232, 'duration': 1.1}, {'end': 175.094, 'text': "And so here's an example right?", 'start': 173.832, 'duration': 1.262}, {'end': 183.003, 'text': 'And I have different options, different pricings, a thousand USD, even more for enterprise hospitals, bigger medical institutions.', 'start': 175.134, 'duration': 7.869}, {'end': 184.785, 'text': 'So you can see how this could work.', 'start': 183.023, 'duration': 1.762}, {'end': 192.82, 'text': "So I'm gonna show you how I did this, how I built this and how I would plan to grow this, okay? So it's a 10-step process.", 'start': 184.825, 'duration': 7.995}], 'summary': 'Automated pneumonia diagnosis to aid doctors and patients, priced at $1000 for regular use and more for enterprise hospitals, with a 10-step growth plan.', 'duration': 32.371, 'max_score': 160.449, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V4160449.jpg'}, {'end': 282.893, 'src': 'embed', 'start': 256.517, 'weight': 5, 'content': [{'end': 261.24, 'text': 'write down what your strengths are, write down what your passions are, write down what the problems you care about are,', 'start': 256.517, 'duration': 4.723}, {'end': 262.982, 'text': 'and then look for the intersection of those three.', 'start': 261.24, 'duration': 1.742}, {'end': 268.025, 'text': 'And in fact, the Japanese have codified this more than 100 years ago with the concept of ikigai.', 'start': 263.262, 'duration': 4.763}, {'end': 275.529, 'text': "It literally means a reason for being, and this chart here shows you what all of that entails what you love, what you're good at,", 'start': 268.505, 'duration': 7.024}, {'end': 277.71, 'text': 'what you can be paid for and what the world needs.', 'start': 275.529, 'duration': 2.181}, {'end': 282.893, 'text': 'and at the intersection of all of those four things is your ikigai, and you wanna find that right?', 'start': 277.71, 'duration': 5.183}], 'summary': "Identify strengths, passions, and problems to find your ikigai, the intersection of what you love, what you're good at, what you can be paid for, and what the world needs.", 'duration': 26.376, 'max_score': 256.517, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V4256517.jpg'}, {'end': 352.758, 'src': 'embed', 'start': 317.871, 'weight': 4, 'content': [{'end': 320.172, 'text': 'How many people have had this problem?', 'start': 317.871, 'duration': 2.301}, {'end': 322.114, 'text': 'How many other ways are there to solve it?', 'start': 320.533, 'duration': 1.581}, {'end': 324.516, 'text': 'What would it cost to offer a solution?', 'start': 322.534, 'duration': 1.982}, {'end': 327.098, 'text': 'And how does the cost compare to the pain of the problem?', 'start': 324.796, 'duration': 2.302}, {'end': 332.503, 'text': "And so, for me, I'm gonna pick one problem in particular of my three passions healthcare, right?", 'start': 327.258, 'duration': 5.245}, {'end': 333.824, 'text': "And so it's gonna be in healthcare.", 'start': 332.583, 'duration': 1.241}, {'end': 337.527, 'text': 'And specifically, I know that image classification technology is really good.', 'start': 334.124, 'duration': 3.403}, {'end': 342.65, 'text': 'So what happens is we have to pick one of those subfields of healthcare.', 'start': 338.748, 'duration': 3.902}, {'end': 345.292, 'text': "I'm gonna say radiology, specifically for lungs.", 'start': 342.931, 'duration': 2.361}, {'end': 352.758, 'text': "That's gonna be the one that I'm most passionate about, because in this example, I have somebody who I know, who's been affected by it,", 'start': 345.593, 'duration': 7.165}], 'summary': 'Focusing on healthcare, considering radiology for lung image classification technology to solve a specific problem.', 'duration': 34.887, 'max_score': 317.871, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V4317871.jpg'}, {'end': 387.703, 'src': 'embed', 'start': 361.132, 'weight': 6, 'content': [{'end': 365.273, 'text': 'Just leave wherever you are for a month and just go to a totally different culture.', 'start': 361.132, 'duration': 4.141}, {'end': 367.594, 'text': 'And you gotta increase your training data right?', 'start': 365.293, 'duration': 2.301}, {'end': 373.296, 'text': 'So find new training data, find new cultures, new values, new languages, new people, new perspectives.', 'start': 367.634, 'duration': 5.662}, {'end': 377.697, 'text': 'And I promise you, you will find a problem that you are passionate about.', 'start': 373.396, 'duration': 4.301}, {'end': 379.418, 'text': 'You will find your ikigai.', 'start': 377.977, 'duration': 1.441}, {'end': 387.703, 'text': 'You need to get out of your bubble and go forth into Strange territory that scares you right? I lived in Amsterdam for a year.', 'start': 379.638, 'duration': 8.065}], 'summary': 'Immerse in a new culture for a month to find passion and ikigai.', 'duration': 26.571, 'max_score': 361.132, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V4361132.jpg'}], 'start': 45.591, 'title': 'Smart medscan and finding ikigai in healthcare', 'summary': "Introduces smart medscan, a tool for medical image classification, aiming to reduce misdiagnoses and improve patient care. it also outlines a 10-step process for identifying healthcare problems and leveraging passions and strengths to find 'ikigai' in the field of image classification for lungs.", 'chapters': [{'end': 175.094, 'start': 45.591, 'title': 'Smart medscan: automating medical image classification', 'summary': 'Introduces smart medscan, a tool for doctors to classify medical images using deep learning and machine learning technologies, aiming to address the shortage of doctors, reduce misdiagnoses, and improve patient care by automating the classification process and providing quick, accurate results.', 'duration': 129.503, 'highlights': ['Smart MedScan automates medical image classification using deep learning and machine learning technologies, addressing the shortage of doctors and reducing misdiagnoses. The tool aims to help doctors classify medical images, specifically x-ray scans for lungs, using deep learning and machine learning technologies to address the shortage of doctors and reduce misdiagnoses.', 'Payment process involves a $10 charge for a single classification, providing quick and accurate results for medical image classification. The payment process involves a $10 charge for a single classification, offering a quick and accurate classification result for medical images.', 'The tool successfully classified a medical image as an example of pneumonia, demonstrating the effectiveness of automation in medical image classification. The tool successfully classified a medical image as an example of pneumonia, showcasing the effectiveness of automating the classification process and its potential to improve patient care.']}, {'end': 397.332, 'start': 175.134, 'title': 'Finding your ikigai in healthcare', 'summary': "Outlines a 10-step process to identify a problem in healthcare, specifically in the field of image classification for lungs, as the speaker shares insights on leveraging passions, strengths, and industry needs to find one's 'ikigai' and emphasizes the importance of exploring different cultures to find a problem to be passionate about.", 'duration': 222.198, 'highlights': ['The importance of identifying a problem in healthcare, particularly in image classification for lungs, using a 10-step process, including market research and creating a business plan.', "The concept of 'ikigai' is introduced as a way to align passions, strengths, industry needs, and societal needs, emphasizing the significance of finding the intersection of these elements.", 'The speaker advocates for exploring different cultures to discover new problems to be passionate about, highlighting the value of broadening perspectives and increasing training data.']}], 'duration': 351.741, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V445591.jpg', 'highlights': ['Smart MedScan automates medical image classification using deep learning and machine learning technologies, addressing the shortage of doctors and reducing misdiagnoses.', 'The tool aims to help doctors classify medical images, specifically x-ray scans for lungs, using deep learning and machine learning technologies to address the shortage of doctors and reduce misdiagnoses.', 'The payment process involves a $10 charge for a single classification, offering a quick and accurate classification result for medical images.', 'The tool successfully classified a medical image as an example of pneumonia, showcasing the effectiveness of automating the classification process and its potential to improve patient care.', 'The importance of identifying a problem in healthcare, particularly in image classification for lungs, using a 10-step process, including market research and creating a business plan.', "The concept of 'ikigai' is introduced as a way to align passions, strengths, industry needs, and societal needs, emphasizing the significance of finding the intersection of these elements.", 'The speaker advocates for exploring different cultures to discover new problems to be passionate about, highlighting the value of broadening perspectives and increasing training data.']}, {'end': 715.063, 'segs': [{'end': 458.853, 'src': 'embed', 'start': 397.372, 'weight': 0, 'content': [{'end': 405.963, 'text': "I have the problem health care for doctors We're gonna automate health care via a classifier and Now, the next step is to do some market research.", 'start': 397.372, 'duration': 8.591}, {'end': 411.045, 'text': 'What are the competing products here? By the way, AI can do literally so much.', 'start': 406.003, 'duration': 5.042}, {'end': 413.125, 'text': 'This is just a subset of what AI can do.', 'start': 411.105, 'duration': 2.02}, {'end': 415.486, 'text': 'It can do anything with natural language processing.', 'start': 413.145, 'duration': 2.341}, {'end': 416.667, 'text': 'It could generate text.', 'start': 415.506, 'duration': 1.161}, {'end': 418.427, 'text': 'It could classify text.', 'start': 417.067, 'duration': 1.36}, {'end': 419.968, 'text': 'It could summarize text.', 'start': 418.667, 'duration': 1.301}, {'end': 423.569, 'text': 'You can use computer vision to see things and classify things.', 'start': 420.768, 'duration': 2.801}, {'end': 430.733, 'text': 'You can generate new types of images for fashion, lines and you can retrieve information in a very intelligent way.', 'start': 423.609, 'duration': 7.124}, {'end': 432.574, 'text': "There's so much information out there.", 'start': 431.133, 'duration': 1.441}, {'end': 439.198, 'text': 'How do you compress it and how do you present to a customer or user the most useful information? That is an AI problem.', 'start': 432.754, 'duration': 6.444}, {'end': 441.58, 'text': "How do you filter what they want and what they don't want?", 'start': 439.218, 'duration': 2.362}, {'end': 447.083, 'text': 'How do you predict the future in finance, in asset prediction, in portfolio optimization?', 'start': 441.92, 'duration': 5.163}, {'end': 451.006, 'text': 'How do you make the best decisions when you have a giant company, a giant organization?', 'start': 447.384, 'duration': 3.622}, {'end': 452.547, 'text': "That's decision analytics.", 'start': 451.286, 'duration': 1.261}, {'end': 454.85, 'text': 'How do you improve the supply chain process?', 'start': 452.868, 'duration': 1.982}, {'end': 458.853, 'text': 'How do you classify these plants that need to be classified?', 'start': 455.37, 'duration': 3.483}], 'summary': 'Automate healthcare with ai for better decision-making and customer satisfaction.', 'duration': 61.481, 'max_score': 397.372, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V4397372.jpg'}, {'end': 528.8, 'src': 'embed', 'start': 499.801, 'weight': 1, 'content': [{'end': 501.242, 'text': "It's going to be a medical image.", 'start': 499.801, 'duration': 1.441}, {'end': 505.724, 'text': "We have our classifier that we're going to build, and it's gonna output what the disease is.", 'start': 501.522, 'duration': 4.202}, {'end': 509.746, 'text': 'Very simple pattern recognition, supervised learning, find a labeled data set.', 'start': 506.064, 'duration': 3.682}, {'end': 511.187, 'text': 'Easy stuff, we can do that.', 'start': 509.986, 'duration': 1.201}, {'end': 515.59, 'text': 'now, if we look at the competitor landscape for this and i found this image,', 'start': 511.787, 'duration': 3.803}, {'end': 519.032, 'text': "you know out there there's general imaging and these people try to do everything.", 'start': 515.59, 'duration': 3.442}, {'end': 525.537, 'text': "you know they're classifying breast cancer, they're classifying, you know, um, all sorts of the whole body.", 'start': 519.032, 'duration': 6.505}, {'end': 528.8, 'text': "then there's like specific niches here.", 'start': 525.537, 'duration': 3.263}], 'summary': 'Building a medical image classifier for disease detection using pattern recognition and supervised learning.', 'duration': 28.999, 'max_score': 499.801, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V4499801.jpg'}, {'end': 613.04, 'src': 'embed', 'start': 582.302, 'weight': 2, 'content': [{'end': 583.463, 'text': "They're trying to do everything.", 'start': 582.302, 'duration': 1.161}, {'end': 588.226, 'text': "What we're going to do is we're going to laser focus on one type of medical image.", 'start': 583.743, 'duration': 4.483}, {'end': 591.049, 'text': 'And those are the lungs, right? lung images.', 'start': 588.246, 'duration': 2.803}, {'end': 592.071, 'text': "We're gonna focus on those.", 'start': 591.089, 'duration': 0.982}, {'end': 593.833, 'text': "We're gonna do that really, really well.", 'start': 592.251, 'duration': 1.582}, {'end': 599.24, 'text': "We're gonna do that better than anybody else because we have the data and the technology to do that freely available.", 'start': 594.033, 'duration': 5.207}, {'end': 607.857, 'text': "when we do that and we build a customer base, then we're going to grow and then we're going to start applying it to all sorts of medical images,", 'start': 600.353, 'duration': 7.504}, {'end': 609.418, 'text': "and that's that's how that looks all right.", 'start': 607.857, 'duration': 1.561}, {'end': 613.04, 'text': 'so we have done our market research and assume that i look through all of these right.', 'start': 609.418, 'duration': 3.622}], 'summary': 'Focusing on lung images, aiming to excel with available data and technology, expanding to other medical images later.', 'duration': 30.738, 'max_score': 582.302, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V4582302.jpg'}, {'end': 695.656, 'src': 'heatmap', 'start': 636.198, 'weight': 5, 'content': [{'end': 637.999, 'text': 'Now the next step is to create a landing page.', 'start': 636.198, 'duration': 1.801}, {'end': 640.761, 'text': 'And to do that, we need a logo, right? We need a logo.', 'start': 638.019, 'duration': 2.742}, {'end': 644.984, 'text': 'Now I use this AI tool to generate a logo for me.', 'start': 641.782, 'duration': 3.202}, {'end': 646.485, 'text': "It's pretty amazing.", 'start': 645.484, 'duration': 1.001}, {'end': 647.526, 'text': "It's called BrandMark.", 'start': 646.525, 'duration': 1.001}, {'end': 650.868, 'text': 'Okay, and what it does is it uses AI to create a logo for you.', 'start': 648.046, 'duration': 2.822}, {'end': 651.628, 'text': 'It generates it.', 'start': 650.928, 'duration': 0.7}, {'end': 653.77, 'text': "It's a generative network that they're using in the back.", 'start': 651.688, 'duration': 2.082}, {'end': 654.691, 'text': 'background here.', 'start': 654.19, 'duration': 0.501}, {'end': 656.312, 'text': "So I'll say it's called Smart MedScan.", 'start': 654.711, 'duration': 1.601}, {'end': 658.894, 'text': 'We bought the GoDaddy domain or whatever domain you buy it from.', 'start': 656.332, 'duration': 2.562}, {'end': 666.08, 'text': "And it's going to be related to hospitals and they're using NLP here to relate images via some semantic similarity index here.", 'start': 659.454, 'duration': 6.626}, {'end': 667.821, 'text': 'Hospital, deep learning.', 'start': 666.46, 'duration': 1.361}, {'end': 668.802, 'text': 'These are my keywords.', 'start': 667.961, 'duration': 0.841}, {'end': 671.024, 'text': "And it's going to then deep learning AI.", 'start': 669.362, 'duration': 1.662}, {'end': 672.245, 'text': 'I see already new AI.', 'start': 671.124, 'duration': 1.121}, {'end': 675.187, 'text': "And I'll say here's the color scheme that we want.", 'start': 672.845, 'duration': 2.342}, {'end': 677.329, 'text': "And it's going to generate a logo for us.", 'start': 675.647, 'duration': 1.682}, {'end': 678.73, 'text': "We don't have to do that ourselves.", 'start': 677.709, 'duration': 1.021}, {'end': 682.713, 'text': "And we're going to use this logo and this color scheme to create a landing page.", 'start': 678.75, 'duration': 3.963}, {'end': 688.293, 'text': 'Right So Check out all these beautiful logos, smart med scan right here, smart med scan.', 'start': 682.793, 'duration': 5.5}, {'end': 694.255, 'text': 'So we would just pick one and we would say, okay, so here it is, smart med scan, deep learning for hospitals.', 'start': 688.313, 'duration': 5.942}, {'end': 695.656, 'text': "Look, it's got the guy with the shirt.", 'start': 694.275, 'duration': 1.381}], 'summary': 'Using ai tool brandmark to generate logo for smart medscan related to hospitals and deep learning, for landing page creation.', 'duration': 59.458, 'max_score': 636.198, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V4636198.jpg'}], 'start': 397.372, 'title': 'Ai in healthcare', 'summary': 'Covers automating healthcare market research through ai capabilities like natural language processing, computer vision, and decision analytics, and discusses leveraging ai in healthcare startup strategy, with a focus on lung imaging and global market targeting.', 'chapters': [{'end': 458.853, 'start': 397.372, 'title': 'Automating healthcare market research', 'summary': 'Discusses automating healthcare through a classifier and the wide-ranging capabilities of ai, including natural language processing, computer vision, decision analytics, and supply chain process improvement.', 'duration': 61.481, 'highlights': ['AI capabilities include natural language processing, computer vision, decision analytics, and supply chain process improvement.', 'AI can generate, classify, and summarize text, as well as retrieve information intelligently.', 'The chapter emphasizes the need to compress and present useful information to users and filter their preferences.', 'The discussion also covers predicting the future in finance, asset prediction, portfolio optimization, and plant classification.']}, {'end': 715.063, 'start': 458.853, 'title': 'Ai in healthcare startup strategy', 'summary': 'Discusses leveraging ai for healthcare, focusing on lung imaging, targeting global markets, using ai tool for logo creation, and creating a landing page, to disrupt the general imaging landscape.', 'duration': 256.21, 'highlights': ['Leveraging AI for healthcare, focusing on lung imaging The chapter emphasizes the potential of AI in healthcare, particularly in the field of lung imaging, to identify diseases through simple pattern recognition and supervised learning.', 'Targeting global markets and disrupting the general imaging landscape The strategy involves targeting markets beyond the US, UK, or EU, and disrupting the general imaging landscape by laser focusing on lung imaging, aiming to do it better than others and eventually expanding to other medical images.', "Using AI tool for logo creation and creating a landing page The AI tool BrandMark is used to generate a logo for 'Smart MedScan', related to hospitals and deep learning, with a color scheme specified, followed by the creation of a landing page to establish the startup's online presence."]}], 'duration': 317.691, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V4397372.jpg', 'highlights': ['AI capabilities include natural language processing, computer vision, decision analytics, and supply chain process improvement.', 'The chapter emphasizes the potential of AI in healthcare, particularly in the field of lung imaging, to identify diseases through simple pattern recognition and supervised learning.', 'The strategy involves targeting markets beyond the US, UK, or EU, and disrupting the general imaging landscape by laser focusing on lung imaging, aiming to do it better than others and eventually expanding to other medical images.', 'The discussion also covers predicting the future in finance, asset prediction, portfolio optimization, and plant classification.', 'The chapter emphasizes the need to compress and present useful information to users and filter their preferences.', "The AI tool BrandMark is used to generate a logo for 'Smart MedScan', related to hospitals and deep learning, with a color scheme specified, followed by the creation of a landing page to establish the startup's online presence.", 'AI can generate, classify, and summarize text, as well as retrieve information intelligently.']}, {'end': 1018.006, 'segs': [{'end': 801.925, 'src': 'embed', 'start': 755.442, 'weight': 0, 'content': [{'end': 759.325, 'text': "we're gonna put the price and then we're gonna say here it is if you're interested, sign up.", 'start': 755.442, 'duration': 3.883}, {'end': 762.788, 'text': "And what we're gonna do is we're gonna find a bunch of medical facilities near us.", 'start': 759.385, 'duration': 3.403}, {'end': 764.93, 'text': "we're gonna ping them and say hey, are you interested?", 'start': 762.788, 'duration': 2.142}, {'end': 767.192, 'text': 'And hopefully we get some people to sign up.', 'start': 765.01, 'duration': 2.182}, {'end': 770.115, 'text': 'And if we do, that is called market validation.', 'start': 767.252, 'duration': 2.863}, {'end': 773.558, 'text': "That's how we know that people actually want our product.", 'start': 770.175, 'duration': 3.383}, {'end': 778.262, 'text': 'And the easiest way to do this is to use MailChimp, which is a tool that I generally use.', 'start': 773.938, 'duration': 4.324}, {'end': 781.664, 'text': 'And so we can say, MailChimp landing page.', 'start': 778.282, 'duration': 3.382}, {'end': 787.109, 'text': "And what MailChimp allows you to do is it's got this free landing page builder, you can try it out.", 'start': 782.384, 'duration': 4.725}, {'end': 792.675, 'text': "And I'll say it's called test, select list, whatever, you know, whatever.", 'start': 788.511, 'duration': 4.164}, {'end': 795.018, 'text': "And we'll begin that.", 'start': 793.376, 'duration': 1.642}, {'end': 799.042, 'text': "And I'll pick one of these.", 'start': 797.36, 'duration': 1.682}, {'end': 801.925, 'text': "Let's just pick this one.", 'start': 800.403, 'duration': 1.522}], 'summary': 'Using mailchimp to create a landing page for market validation of a medical product.', 'duration': 46.483, 'max_score': 755.442, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V4755442.jpg'}, {'end': 899.447, 'src': 'embed', 'start': 867.921, 'weight': 4, 'content': [{'end': 873.585, 'text': "So let's share that landing page so we can share it on our social networks, you know, on Facebook, whatever it is like.", 'start': 867.921, 'duration': 5.664}, {'end': 875.527, 'text': 'hey, sign up for this.', 'start': 873.585, 'duration': 1.942}, {'end': 879.57, 'text': 'Hey, sign up, you know, make it a better pitch.', 'start': 875.587, 'duration': 3.983}, {'end': 885.975, 'text': "But like, hey, you know, if you're interested, I have a startup for doctors.", 'start': 879.61, 'duration': 6.365}, {'end': 886.995, 'text': 'Sign up here.', 'start': 886.315, 'duration': 0.68}, {'end': 887.716, 'text': 'Sign up here.', 'start': 887.216, 'duration': 0.5}, {'end': 888.356, 'text': 'Coming soon.', 'start': 887.776, 'duration': 0.58}, {'end': 889.197, 'text': 'Right Whatever it is.', 'start': 888.416, 'duration': 0.781}, {'end': 890.418, 'text': 'And then, you know, add the link.', 'start': 889.237, 'duration': 1.181}, {'end': 899.447, 'text': 'And then the other thing we can do is we can go to Maps and then we can say inside of Google Maps what are the places that are near me?', 'start': 891.223, 'duration': 8.224}], 'summary': 'Discussion on sharing landing page on social networks and using google maps to find nearby places.', 'duration': 31.526, 'max_score': 867.921, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V4867921.jpg'}, {'end': 1025.069, 'src': 'embed', 'start': 994.555, 'weight': 1, 'content': [{'end': 996.355, 'text': 'And this is called The Lean Startup Mentality.', 'start': 994.555, 'duration': 1.8}, {'end': 998.316, 'text': 'I highly recommend checking out this book.', 'start': 996.415, 'duration': 1.901}, {'end': 1002.678, 'text': "It's a great book, it's by Eric Ries, and it talks about how to do this.", 'start': 999.277, 'duration': 3.401}, {'end': 1006.7, 'text': "Basically, TLDR of the entire book is don't just try to raise money.", 'start': 1002.798, 'duration': 3.902}, {'end': 1013.424, 'text': 'Just build as much as you can without while being frugal, while being frugal.', 'start': 1006.8, 'duration': 6.624}, {'end': 1015.805, 'text': 'And that is how you build a real business.', 'start': 1013.544, 'duration': 2.261}, {'end': 1018.006, 'text': 'Okay, so now onto step seven.', 'start': 1016.445, 'duration': 1.561}, {'end': 1020.908, 'text': 'Step seven is going to be to design a pipeline.', 'start': 1018.366, 'duration': 2.542}, {'end': 1025.069, 'text': 'So what we know we need is, first of all, we need there to be a database of users.', 'start': 1020.968, 'duration': 4.101}], 'summary': 'The lean startup mentality recommends building without excessive spending, emphasizing frugality to create a real business, and advises designing a user database as a step in the process.', 'duration': 30.514, 'max_score': 994.555, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V4994555.jpg'}], 'start': 715.063, 'title': 'Creating landing pages and lean startup mentality', 'summary': 'Covers strategies for creating landing pages to validate market interest, using mailchimp, and applying lean startup mentality to gain 25 signups, convert them to paid clients, collect feedback, and eventually hire a full stack engineer and a salesman.', 'chapters': [{'end': 773.558, 'start': 715.063, 'title': 'Creating landing page for market validation', 'summary': 'Discusses the strategy of creating a landing page with a logo, email field, product description, and price to validate the market interest in the product by reaching out to medical facilities.', 'duration': 58.495, 'highlights': ['Creating a landing page with a logo, email field, product description, and price to validate market interest.', 'The goal is to find out if there are potential customers before investing time and resources into building the product.', 'Reaching out to medical facilities to gauge their interest and validate the market for the product.', 'Market validation is achieved if people sign up through the landing page, indicating genuine interest in the product.']}, {'end': 1018.006, 'start': 773.938, 'title': 'Mailchimp landing page and lean startup mentality', 'summary': 'Discusses creating a landing page using mailchimp, sharing it on social networks, and the lean startup mentality, aiming to get 25 signups for the landing page, create the app, convert signups to paid clients, collect feedback, and eventually hire a full stack engineer and a salesman.', 'duration': 244.068, 'highlights': ['The chapter explains the process of creating a landing page using MailChimp and emphasizes the importance of linking the domain to the landing page.', 'It discusses sharing the landing page on social networks and reaching out to potential clients using Google Maps, highlighting the strategic approach to gathering potential leads.', 'The chapter outlines the Lean Startup Mentality, aiming to get 25 signups for the landing page, create the app, convert signups to paid clients, collect feedback, and eventually hire a full stack engineer and a salesman.', 'It details the step-by-step plan, including collecting 100 paid clients, repeating feedback collection, and eventual team expansion, showcasing a quantifiable approach to business growth.', 'The chapter concludes with the Lean Startup Mentality approach, emphasizing the importance of frugal building and not just focusing on raising money, promoting a sustainable and strategic business approach.']}], 'duration': 302.943, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V4715063.jpg', 'highlights': ['Market validation is achieved if people sign up through the landing page, indicating genuine interest in the product.', 'The chapter outlines the Lean Startup Mentality, aiming to get 25 signups for the landing page, create the app, convert signups to paid clients, collect feedback, and eventually hire a full stack engineer and a salesman.', 'The chapter explains the process of creating a landing page using MailChimp and emphasizes the importance of linking the domain to the landing page.', 'The goal is to find out if there are potential customers before investing time and resources into building the product.', 'It discusses sharing the landing page on social networks and reaching out to potential clients using Google Maps, highlighting the strategic approach to gathering potential leads.']}, {'end': 1547.843, 'segs': [{'end': 1054.589, 'src': 'embed', 'start': 1018.366, 'weight': 0, 'content': [{'end': 1020.908, 'text': 'Step seven is going to be to design a pipeline.', 'start': 1018.366, 'duration': 2.542}, {'end': 1025.069, 'text': 'So what we know we need is, first of all, we need there to be a database of users.', 'start': 1020.968, 'duration': 4.101}, {'end': 1028.112, 'text': "authentication functionality right, that's very basic.", 'start': 1025.609, 'duration': 2.503}, {'end': 1030.834, 'text': 'just sign up login right, that we know.', 'start': 1028.112, 'duration': 2.722}, {'end': 1031.394, 'text': 'we need that.', 'start': 1030.834, 'duration': 0.56}, {'end': 1034.316, 'text': 'and we also need a database to store those users.', 'start': 1031.394, 'duration': 2.922}, {'end': 1035.457, 'text': "so we're gonna use sequel.", 'start': 1034.316, 'duration': 1.141}, {'end': 1041.621, 'text': 'sequel is the go-to default database that you know we use these days, and obviously there are others.', 'start': 1035.457, 'duration': 6.164}, {'end': 1045.143, 'text': "but for you know, we're trying to build a prototype as fast as possible.", 'start': 1041.621, 'duration': 3.522}, {'end': 1052.288, 'text': "rapid experimentation it's one of our values at school of AI just rapidly experiment until we get to whatever solution we need.", 'start': 1045.143, 'duration': 7.145}, {'end': 1054.589, 'text': 'try, fail, try, fail, try, fail.', 'start': 1052.288, 'duration': 2.301}], 'summary': 'Design a pipeline for user database and authentication, using sequel for rapid prototyping.', 'duration': 36.223, 'max_score': 1018.366, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V41018366.jpg'}, {'end': 1113.495, 'src': 'embed', 'start': 1081.601, 'weight': 4, 'content': [{'end': 1084.282, 'text': "So once a user signs up, they'll need to be able to pay.", 'start': 1081.601, 'duration': 2.681}, {'end': 1089.183, 'text': 'We need to be able to collect their card information, easily store it in a secured way.', 'start': 1084.302, 'duration': 4.881}, {'end': 1094.144, 'text': 'And then once they do that, we want to allow them to upload their images.', 'start': 1089.743, 'duration': 4.401}, {'end': 1096.725, 'text': 'So flask will natively allow us to do that.', 'start': 1094.204, 'duration': 2.521}, {'end': 1101.066, 'text': 'Speaking of flask flask is a web development framework for Python.', 'start': 1096.885, 'duration': 4.181}, {'end': 1104.127, 'text': "And the reason that I chose flask is because I'm most familiar with it.", 'start': 1101.406, 'duration': 2.721}, {'end': 1106.088, 'text': 'However, Django works just as well.', 'start': 1104.447, 'duration': 1.641}, {'end': 1106.948, 'text': 'Pyramid air pair.', 'start': 1106.168, 'duration': 0.78}, {'end': 1113.495, 'text': 'and there are a million other ones, but Flask seems to have a lot of good documentation, a lot of good repositories on GitHub,', 'start': 1107.248, 'duration': 6.247}], 'summary': 'Users need to pay, store card info securely, upload images using flask framework for python.', 'duration': 31.894, 'max_score': 1081.601, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V41081601.jpg'}, {'end': 1218.593, 'src': 'embed', 'start': 1190.824, 'weight': 5, 'content': [{'end': 1194.148, 'text': 'And one that I found worked pretty well was called Flask SaaS.', 'start': 1190.824, 'duration': 3.324}, {'end': 1197.578, 'text': 'which this guy built here.', 'start': 1195.737, 'duration': 1.841}, {'end': 1199.199, 'text': 'And we can build off of his work.', 'start': 1197.618, 'duration': 1.581}, {'end': 1203.102, 'text': "So we're gonna go ahead and clone, download this straight from GitHub.", 'start': 1199.219, 'duration': 3.883}, {'end': 1207.525, 'text': "And we'll get into that in a second.", 'start': 1205.043, 'duration': 2.482}, {'end': 1209.767, 'text': 'And then we wanna add upload functionality.', 'start': 1207.945, 'duration': 1.822}, {'end': 1213.609, 'text': 'Who has upload functionality? Flask Image Uploader.', 'start': 1210.127, 'duration': 3.482}, {'end': 1215.31, 'text': "Oh, he's got a running demo right there.", 'start': 1213.749, 'duration': 1.561}, {'end': 1218.593, 'text': "And he's got the code right here.", 'start': 1215.811, 'duration': 2.782}], 'summary': 'Using flask saas and flask image uploader for building off existing work, including upload functionality.', 'duration': 27.769, 'max_score': 1190.824, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V41190824.jpg'}, {'end': 1318.896, 'src': 'embed', 'start': 1297.128, 'weight': 3, 'content': [{'end': 1308.03, 'text': 'So the whole idea behind transfer learning is to not have us have to waste a lot of our time training our model on an extremely large data set.', 'start': 1297.128, 'duration': 10.902}, {'end': 1311.451, 'text': "It's also not gonna require a lot of computational power.", 'start': 1308.25, 'duration': 3.201}, {'end': 1318.896, 'text': "So, the way to do that is to first decide what technique we're going to use for medical imaging classification.", 'start': 1312.171, 'duration': 6.725}], 'summary': 'Transfer learning reduces training time and computational power for medical imaging classification.', 'duration': 21.768, 'max_score': 1297.128, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V41297128.jpg'}], 'start': 1018.366, 'title': 'Designing data pipelines and building web apps with flask', 'summary': 'Discusses the need for a user database with authentication using sequel, flask for web app development, stripe integration for payments, image upload functionality, transfer learning with keras for medical image classification, and app deployment.', 'chapters': [{'end': 1054.589, 'start': 1018.366, 'title': 'Designing a data pipeline', 'summary': 'Discusses the need for a user database with authentication functionality using sequel for rapid experimentation and prototype development at the school of ai.', 'duration': 36.223, 'highlights': ['Sequel is the default database for rapid experimentation and prototype development at the School of AI, emphasizing the value of try, fail, try, fail, try, fail.', 'The need for a database of users with authentication functionality for sign up and login is identified as a basic requirement.', 'Emphasizing the importance of rapid experimentation and the value of try, fail, try, fail, try, fail for reaching the desired solution.']}, {'end': 1547.843, 'start': 1054.589, 'title': 'Building a web app with flask for medical image diagnosis', 'summary': 'Discusses using flask for user authentication and database storage, integrating stripe for payments, adding image upload functionality, leveraging pre-existing github repositories for app components, performing transfer learning with keras for medical image classification, and deploying the web app.', 'duration': 493.254, 'highlights': ['Performing transfer learning using pre-trained models like Inception for medical imaging classification to improve accuracy and reduce training time. The technique of transfer learning is employed to leverage pre-trained models like Inception, cutting off the final layer and retraining it on a medical imaging dataset to improve accuracy and reduce computational power.', 'Utilizing Flask for user authentication, database storage, and integrating Stripe for payment functionality. The use of Flask is discussed for user authentication, database storage, and integrating Stripe for payment functionality, allowing users to securely store card information and make payments.', 'Leveraging pre-existing GitHub repositories for web app components, such as Flask SaaS for user authentication, payments, and database, and Flask Image Uploader for image upload functionality. The chapter emphasizes the use of pre-existing GitHub repositories like Flask SaaS for user authentication, payments, and database, and Flask Image Uploader for image upload functionality, enabling rapid prototyping and component integration.']}], 'duration': 529.477, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V41018366.jpg', 'highlights': ['Sequel is the default database for rapid experimentation and prototype development at the School of AI, emphasizing the value of try, fail, try, fail, try, fail.', 'The need for a database of users with authentication functionality for sign up and login is identified as a basic requirement.', 'Emphasizing the importance of rapid experimentation and the value of try, fail, try, fail, try, fail for reaching the desired solution.', 'Performing transfer learning using pre-trained models like Inception for medical imaging classification to improve accuracy and reduce training time.', 'Utilizing Flask for user authentication, database storage, and integrating Stripe for payment functionality.', 'Leveraging pre-existing GitHub repositories for web app components, such as Flask SaaS for user authentication, payments, and database, and Flask Image Uploader for image upload functionality.']}, {'end': 2371.468, 'segs': [{'end': 1617.768, 'src': 'heatmap', 'start': 1568.948, 'weight': 0, 'content': [{'end': 1570.089, 'text': "So that's transfer learning.", 'start': 1568.948, 'duration': 1.141}, {'end': 1572.67, 'text': "Now we're onto step nine where we're almost done.", 'start': 1570.129, 'duration': 2.541}, {'end': 1573.871, 'text': 'We got to write our web app.', 'start': 1572.69, 'duration': 1.181}, {'end': 1578.533, 'text': "So let's write our web app, starting with that boilerplate that we just downloaded right here.", 'start': 1573.891, 'duration': 4.642}, {'end': 1580.334, 'text': "It's called flask SAS master.", 'start': 1578.613, 'duration': 1.721}, {'end': 1584.096, 'text': "Okay So it's called flask SAS master one.", 'start': 1580.354, 'duration': 3.742}, {'end': 1585.176, 'text': "That's it.", 'start': 1584.696, 'duration': 0.48}, {'end': 1586.237, 'text': "So I'm going to copy that.", 'start': 1585.196, 'duration': 1.041}, {'end': 1587.657, 'text': "I'm going to go here.", 'start': 1586.757, 'duration': 0.9}, {'end': 1593.2, 'text': "I'm going to then let's see, let's make this bigger.", 'start': 1587.677, 'duration': 5.523}, {'end': 1594.101, 'text': "Let's do this.", 'start': 1593.56, 'duration': 0.541}, {'end': 1595.421, 'text': 'I love terminal.', 'start': 1594.701, 'duration': 0.72}, {'end': 1596.182, 'text': 'Here we go.', 'start': 1595.882, 'duration': 0.3}, {'end': 1596.922, 'text': 'CD into it.', 'start': 1596.302, 'duration': 0.62}, {'end': 1598.483, 'text': "Boom We're there.", 'start': 1597.622, 'duration': 0.861}, {'end': 1600.297, 'text': 'Okay, Sublime, open it up.', 'start': 1598.996, 'duration': 1.301}, {'end': 1601.298, 'text': "Here's our file.", 'start': 1600.617, 'duration': 0.681}, {'end': 1610.744, 'text': "Now, where were we? What was the name of that file? It's called flasksass.", 'start': 1601.938, 'duration': 8.806}, {'end': 1615.847, 'text': 'And what are the instructions to run it? Setup, make install and make dev.', 'start': 1611.204, 'duration': 4.643}, {'end': 1617.768, 'text': 'Great We can do that.', 'start': 1616.227, 'duration': 1.541}], 'summary': 'Implementing web app using flask sas master for transfer learning.', 'duration': 48.82, 'max_score': 1568.948, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V41568948.jpg'}, {'end': 1730.874, 'src': 'embed', 'start': 1701.512, 'weight': 2, 'content': [{'end': 1703.333, 'text': "Let's look at what this app has in here.", 'start': 1701.512, 'duration': 1.821}, {'end': 1705.894, 'text': 'How is this app structured? We need to understand it.', 'start': 1703.633, 'duration': 2.261}, {'end': 1708.335, 'text': "So let's look in the config file.", 'start': 1706.274, 'duration': 2.061}, {'end': 1715.06, 'text': 'So in the config file, this person added the credentials for the database to SQLite database.', 'start': 1708.695, 'duration': 6.365}, {'end': 1718.123, 'text': "They've got the mail server, great stuff like that.", 'start': 1715.841, 'duration': 2.282}, {'end': 1724.568, 'text': 'What we want to do is we want to add an upload folder, right? So these are global keywords.', 'start': 1718.663, 'duration': 5.905}, {'end': 1730.874, 'text': "So one thing we want to do is we want to add an upload folder that's going to store whatever images that doctors or whoever uploads.", 'start': 1724.648, 'duration': 6.226}], 'summary': 'App config includes sqlite database credentials and mail server settings. plan to add upload folder for images.', 'duration': 29.362, 'max_score': 1701.512, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V41701512.jpg'}, {'end': 2290.438, 'src': 'embed', 'start': 2260.168, 'weight': 3, 'content': [{'end': 2262.948, 'text': "And there's a lot more we can do, right? So that was this 10th step.", 'start': 2260.168, 'duration': 2.78}, {'end': 2270.77, 'text': 'Now in the future, what can we do? Well, we can obviously add a mobile app so doctors can classify with their mobile devices, with their cameras.', 'start': 2263.389, 'duration': 7.381}, {'end': 2277.092, 'text': 'And we can have it happen on the device with low power inferencing techniques, CoreML for iOS.', 'start': 2271.17, 'duration': 5.922}, {'end': 2280.953, 'text': 'And Android has a whole suite dedicated to this.', 'start': 2277.392, 'duration': 3.561}, {'end': 2283.894, 'text': 'then we can also make a better model right.', 'start': 2281.793, 'duration': 2.101}, {'end': 2290.438, 'text': 'so one thing we can do is we can add batch inference so they can upload many, many images all at once, not just one.', 'start': 2283.894, 'duration': 6.544}], 'summary': 'Plan to add mobile app for doctors with low power inferencing techniques and batch inference for multiple image uploads.', 'duration': 30.27, 'max_score': 2260.168, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V42260168.jpg'}, {'end': 2347.172, 'src': 'embed', 'start': 2320.572, 'weight': 1, 'content': [{'end': 2324.134, 'text': "So it's this continuous training pipeline where it's improving over time.", 'start': 2320.572, 'duration': 3.562}, {'end': 2327.136, 'text': 'learning from its users data, from your users data.', 'start': 2324.654, 'duration': 2.482}, {'end': 2330.459, 'text': "And this is very, very powerful and it's what we want to do.", 'start': 2327.436, 'duration': 3.023}, {'end': 2333.661, 'text': 'We want to make our system better with every image that it sees.', 'start': 2330.499, 'duration': 3.162}, {'end': 2338.045, 'text': 'We want it to continuously improve and we want it to be robust and to always stay online.', 'start': 2333.961, 'duration': 4.084}, {'end': 2347.172, 'text': 'So, obviously I did not cover every single thing needed to build an AI startup, but hopefully I gave you some education, some inspiration,', 'start': 2338.205, 'duration': 8.967}], 'summary': 'Continuous learning pipeline for improving ai with user data, aiming for robustness and online availability.', 'duration': 26.6, 'max_score': 2320.572, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V42320572.jpg'}], 'start': 1547.843, 'title': 'Web app development and ai deployment', 'summary': 'Covers transfer learning, web app development with flask, database setup, authentication, and app structure. it also details stripe payment integration, ai system deployment, continuous serving pipeline, and potential future additions like a mobile app and batch inference functionality.', 'chapters': [{'end': 1882.411, 'start': 1547.843, 'title': 'Building web app with flask', 'summary': 'Covers the process of transfer learning and building a web app using flask, including setting up the database, authentication, and app structure.', 'duration': 334.568, 'highlights': ['The process of transfer learning and building a web app using Flask is explained. The chapter covers the process of transfer learning and building a web app using Flask, including setting up the database, authentication, and app structure.', 'The steps for setting up the database and running the application are described. The steps for setting up the database, creating a SQL database, and running the application with Python 3 are outlined.', 'Instructions for modifying and structuring the web app, including adding an upload folder and customizing the appearance, are provided. Instructions for modifying and structuring the web app are provided, including adding an upload folder for storing images, customizing the appearance with HTML and CSS, and adding a logo.']}, {'end': 2371.468, 'start': 1882.431, 'title': 'Stripe payment integration and ai deployment', 'summary': 'Details the integration of payment functionality using stripe api and the deployment of an ai system, including building a continuous serving pipeline for model improvement, with a mention of possible future additions such as a mobile app for doctors and batch inference functionality.', 'duration': 489.037, 'highlights': ["The integration of payment functionality using Stripe API The speaker explains the process of creating a Stripe account, obtaining API keys, and integrating code snippets provided by Stripe to facilitate card payments, with a mention of Stripe's business model of charging a cut.", 'Deployment of an AI system with continuous serving pipeline for model improvement The transcript outlines the process of deploying an AI system using Heroku, followed by a discussion on building a continuous serving pipeline for model improvement, emphasizing the use of TensorFlow serving for real-time model updating and improvement.', 'Future additions such as a mobile app for doctors and batch inference functionality The speaker mentions potential future additions to the system, including the development of a mobile app for doctors to classify images, and the incorporation of batch inference functionality, along with the suggestion to explore CoreML for iOS and dedicated tools for Android.']}], 'duration': 823.625, 'thumbnail': 'https://coursnap.oss-ap-southeast-1.aliyuncs.com/video-capture/NzmoPqte4V4/pics/NzmoPqte4V41547843.jpg', 'highlights': ['Covers transfer learning, web app development with Flask, database setup, authentication, and app structure.', 'Integration of payment functionality using Stripe API and deployment of an AI system with continuous serving pipeline for model improvement.', 'Instructions for modifying and structuring the web app, including adding an upload folder and customizing the appearance, are provided.', 'Future additions such as a mobile app for doctors and batch inference functionality.']}], 'highlights': ['Smart MedScan automates medical image classification using deep learning and machine learning technologies, addressing the shortage of doctors and reducing misdiagnoses.', 'The tool aims to help doctors classify medical images, specifically x-ray scans for lungs, using deep learning and machine learning technologies to address the shortage of doctors and reduce misdiagnoses.', 'The payment process involves a $10 charge for a single classification, offering a quick and accurate classification result for medical images.', 'The tool successfully classified a medical image as an example of pneumonia, showcasing the effectiveness of automating the classification process and its potential to improve patient care.', 'The chapter outlines the Lean Startup Mentality, aiming to get 25 signups for the landing page, create the app, convert signups to paid clients, collect feedback, and eventually hire a full stack engineer and a salesman.', 'The strategy involves targeting markets beyond the US, UK, or EU, and disrupting the general imaging landscape by laser focusing on lung imaging, aiming to do it better than others and eventually expanding to other medical images.', 'AI capabilities include natural language processing, computer vision, decision analytics, and supply chain process improvement.', 'The chapter emphasizes the potential of AI in healthcare, particularly in the field of lung imaging, to identify diseases through simple pattern recognition and supervised learning.', 'Market validation is achieved if people sign up through the landing page, indicating genuine interest in the product.', 'The need for a database of users with authentication functionality for sign up and login is identified as a basic requirement.', 'Integration of payment functionality using Stripe API and deployment of an AI system with continuous serving pipeline for model improvement.', 'Future additions such as a mobile app for doctors and batch inference functionality.']}