DermaSkan

? DermaSkan Chrome Extension Update Coming Soon! ? ?? What's New? We're excited to announce a major update to the DermaSkan tool chrome extension! Our team has been diligently working to enhance the tool by retraining our AI model with fresh data. We've upgraded our classifier and rating system to provide even more nuanced product recommendations, akin to a dermatologist's precise advice. ?? Enhanced Recommendation System: Now, our tool not only rates how well a moisturizer suits you (ranging from Not Recommended to Great and Recommended) but also explains the reasoning behind these ratings. This added transparency helps you understand and trust why a product is deemed a perfect match or not. ? Stay Tuned: In the next two weeks, we're rolling out this new version, featuring an AI that's re-trained to offer even better, more insightful explanations. Haven't tried our tool yet? Stay tuned and get ready to explore a smarter way to choose skincare products! ?? Your Skincare Companion: As we continue to innovate, your feedback is invaluable in shaping DermaSkan into the ultimate skincare companion. Thank you for being a part of our journey! #SkincareTechnology #AIinHealthcare #ChromeExtension #PersonalizedSkincare #TechInnovation #UserExperience #Dermatology

???????????????????? ???????????? ?????????????? ??????? ??????????????: We’ve rolled out exciting enhancements to our Results page! The traditional bar has been replaced with a dynamic meter, making it easier for users to interpret their skincare evaluations at a glance. Additionally, we’ve introduced a new feature that explains DermaSkan’s predictions in detail, helping users understand why a product is mapped as a good or bad match based on their responses. This transparency empowers users to make more informed skincare decisions. ??????????? ?????????????? ??????????????????????: Our data science team is hard at work gathering more accurate and reliable information on product suitability for different skin types. This rigorous approach ensures that our recommendations are more tailored than ever. We're also refining our classifier algorithm, which basically classifies which product is suitable for which skin concern, to improve the accuracy of our product recommendations based on specific skin concerns. ????????????????? ????????????????: As we continue to innovate and improve, your feedback remains invaluable. These updates are just the beginning of our commitment to refining DermaSkan into the ultimate skincare companion. Thank you for being part of our journey! #SkincareInnovation #AIinHealthcare #DataScience #TechDevelopment #UserExperience #ProductFeedback #Dermatology #StartupJourney

? ?????????????????? ???????????? ???????????? ? ?? Launch Success: We're excited to announce that DermaSkan is now live on the Chrome Web Store as of Tuesday, 11:30 AM! ?? Our journey to revolutionize skincare shopping has officially begun. ?? Early Insights: As with any innovative tech, we’ve encountered some early challenges. Some users reported intermittent issues with score loading, occasional extension crashes during high-traffic periods, and unexpected logouts. Our team is actively addressing these to enhance stability and user experience. ?? On the Horizon: Design phase for Version 2 of our AI model has kicked off! With updated insights from dermatologists, we're gearing up to refine our algorithms for even more precise skincare recommendations. ?? Beta Testers Wanted: Got a keen eye for detail and a passion for skincare? Join us as a beta tester and help shape the future of DermaSkan! Download the extension, the link is in the comments below! ? Add the extension to your chrome. ? Navigate to any moisturizer page on Sephora. Click on the DermaSkan chrome extension to initiate the evaluation process. ? Note that DermaSkan only works on moisturizers on the Sephora USA website (we’re working to add products in other skincare categories). ?? Front-End Enhancements: Our development team is making great strides with new features, including: ? An enhanced scoring meter for clearer understanding of product ratings. ? A feedback option directly within the extension, allowing users to comment on their scores and suggest improvements. ? Layout optimizations to ensure a seamless and intuitive user experience across all devices. ?? Your Experience Matters: Your feedback is crucial in helping us fine-tune DermaSkan. Together, we can create a tool that not only meets but exceeds your skincare shopping needs. Thank you for your continued support and enthusiasm as we navigate these early days and move forward with exciting new developments! #SkincareInnovation #TechDevelopment #AIinHealthcare #StartupJourney #Dermatology #DataScience #UserExperience #ProductFeedback

? ?????????????????? ???????????? ???????????? ? ?? Major Milestone Achieved: This week, we're thrilled to announce that we've completed the training and testing of our MVP! After rigorous evaluation against dermatologist-annotated data, we've selected the machine learning model that delivers the most accurate and personalized skincare recommendations. ?? Deep Learning Developments: Concurrently, our data science team is making exciting progress on a deep learning model, enhancing our capabilities to analyze complex skincare needs with even greater precision. ?? Tech Advancements: Our full-stack team has been hard at work, successfully deploying our backend on GCP Cloud Run and setting up storage buckets. Despite initial challenges with bucket access, our team swiftly navigated through these hurdles, ensuring a smooth integration with our MVP model. ?? Countdown to Launch: With the official release of our Chrome extension set for August 5th (Monday), we are in the final stages of preparing for a seamless launch. Our tool is designed to eliminate the guesswork in selecting skincare products by providing clear, dermatologist-backed recommendations. ?? How Our Tool Works: Imagine you're browsing moisturizers on Sephora. Instead of sifting through countless options, our DermaSkan Chrome extension simplifies the process. 1. Just click on a product, activate our extension. 2. Answer a few questions about your skin. 3. Receive a personalized score ranging from 'Not Recommended', ‘Ok’, ‘Great’ to ‘Excellent’. This score is calculated using an AI model that emulates a dermatologist's expert judgment, ensuring you choose the best product for your skin. ?? Why Choose DermaSkan?: Our tool not only reduces ambiguity in product selection but also boosts your confidence in your skincare choices, backed by solid, data-driven insights. Get ready to transform your skincare shopping experience with DermaSkan! Thank you for your continued support as we move closer to transforming the skincare industry with innovative, technology-driven solutions. #SkincareInnovation #AIinHealthcare #PersonalizedSkincare #MachineLearning #DeepLearning #TechDevelopment #StartupJourney #Dermatology #DataScience #ProductLaunch

? ?????????????????? ???????????? ???????????? ? ?? Innovative Approach to Skincare: This week at DermaSkan, we’re excited to highlight the unique aspects of our product as we approach our MVP launch. Built from scratch, our comprehensive product dataset and synthetically generated customer profiles are meticulously crafted with insights from dermatologists. This innovative combination ensures our AI-driven classifier accurately classifies skincare products based on dermatological criteria and our rating algorithm mimics the nuanced way a dermatologist rates products against customer profiles, ensuring precise and personalized skincare recommendations. ?? Data Science Milestones: Our data scientists have commenced exploratory data analysis on our MVP dataset. They are diligently progressing to develop, train, and fine-tune three machine learning models and one advanced deep learning model. These will soon undergo rigorous testing against a dermatologist-created dataset to select the best-performing model for our platform. ??? Technology Integration: Our full-stack team has been testing the prototype model this week, focusing on optimizing the integration by fine-tuning the input encoding and output decoding processes. Their efforts are crucial as we prepare for a seamless launch. ?? Upcoming Launch Plans: As we finalize our best model by 08/01, anticipation builds for the official Chrome extension release scheduled for 08/04. The forthcoming weeks are slated for integrating the MVP model, ensuring everything is polished and primed for an exceptional user experience. ??? Sneak Peek: Here is our exclusive demo that showcases how our prototype model works, offering a glimpse into the personalized skincare solutions we’re creating. Thank you for your continued support as we blend technology with expert dermatological insights to transform the skincare industry. #SkincareInnovation #AIinHealthcare #PersonalizedSkincare #MachineLearning #DeepLearning #TechDevelopment #StartupJourney #Dermatology #DataScience #ProductLaunch

? ?????????????????? ???????????? ???????????? ? ?? ???????????????? ????????????: We are thrilled to announce the completion of our advanced product classifier! This AI-driven tool now effectively determines which skincare products are suitable for different skin types and concerns, and identifies those suitable for sensitive skin, automating the expert analysis traditionally performed by dermatologists. This major milestone brings us closer to the MVP readiness with a robust dataset now featuring 25,000 entries! ?? ?????? ???????????????? ?????????????? ????????????: Our MVP dataset is now fully compiled, featuring customer profiles, their matched products, and detailed product features. Each entry includes a rating from "Not recommended", "Ok", "Great", and "Excellent," reflecting how well a product fits a specific profile, modeled after dermatologist scoring patterns, but automated through our cutting-edge AI algorithms. ?? ?????????????????????????? ????????????: A crucial phase begins as our dermatologists will validate the AI-generated ratings for 150 of these entries to ensure accuracy and relevance. Their insights will guide any necessary refinements to our algorithms, ensuring our model performs at the highest standards. ??? ???????? ????????????????????????: On the tech front, we've upgraded our Chrome extension with new capabilities. It now restricts operations to only those products listed in our database, thanks to our newly developed product verification API. Additionally, our user response API is set to capture and analyze consumer feedback efficiently. ?? ?????????????? ??????????: As our dermatologists validate the rating algorithm, the rest of the team is diving into data preprocessing and analysis to ensure no time is wasted. This parallel processing will accelerate our journey towards an operational MVP. Thank you for your support as we continue to blend advanced technology with dermatological expertise to revolutionize personalized skincare. #SkincareInnovation #AIinHealthcare #PersonalizedSkincare #MachineLearning #DigitalHealth #TechForGood #StartupJourney #Dermatology #DataScience #ProductDevelopment

? ?????????????????? ???????????? ???????????? ? ?? ???????????????? ????????????: Another productive week at DermaSkan as we edge closer to the MVP stage! We've made a significant leap in enhancing our prototype by incorporating additional product information into our dataset. These new features are proving invaluable, greatly enriching our model's capacity to transition smoothly from a prototype to a fully operational MVP. ??? ???????????????????? ????????????: Our product classifier has undergone major developments this week. Originally slated for completion by 07/07, we've extended the deadline to 07/15 to ensure its precision. Upon completion, it will undergo a meticulous review by our dermatologists to verify its accuracy in identifying features like suitability for sensitive skin or specific skin concerns. ?? ?????? ???????????????? ??????????????: We've compiled a comprehensive training dataset for our MVP, which now includes: ? Customer Profiles: Expanded to encompass a wide array of skin types and concerns, including 10 different skin concerns and sensitivity options. ? Product Details: Our catalog now boasts 630 products, each classified for specific skin types, concerns, fragrance presence, and suitability for sensitive skin. ? Enhanced Matching: Post-classification, we'll strategically group products to match them with synthetic customer profiles based on their skin type and concerns, ensuring a tailored user experience. ?? ???????????? ?????????????????? ????????????????????: We've devised a robust rating algorithm that assigns weights to various customer and product attributes, enhancing the relevance and accuracy of our product recommendations through a sophisticated weighted cosine similarity calculation. ?? ???????? ????????????????????????: Exciting advancements on the tech front! We've implemented a new chrome extension capable of scraping product details directly from selected websites and integrating user responses, streamlining the data acquisition process significantly. ?? ?????????????? ??????????: As we proceed to refine our classifier and prepare for the MVP training data preprocessing, stay tuned for more updates. Our commitment to merging cutting-edge technology with skincare continues to drive us towards delivering personalized, science-backed skincare solutions. Thank you for your continued support as we navigate this exciting phase of our journey, pushing the boundaries. #SkincareInnovation #AIinHealthcare #PersonalizedSkincare #MachineLearning #DigitalHealth #TechForGood #StartupJourney #Dermatology #DataScience #ProductDevelopment

? ?????????????????? ???????????? ???????????? ? ?? ???????????????? ????????????: This week, we're excited to report substantial progress towards our MVP. We've enriched our prototype with a newly curated dataset, which includes 32 additional, dermatologist-reviewed test cases. This dataset enhances our prototype's ability to predict the suitability of skincare products for individual users, utilizing a precise 4-point rating scale that ranges from "not recommended" to "excellent." ?? ???????????????????? ????????????: Significant strides have been made in our ingredient classifier project. We've successfully completed the initial phases of identifying and cataloging key ingredients that address specific skin concerns from product descriptions. We are now advancing to the algorithm development phase, which is poised to intelligently classify skincare products based on these ingredients, with a target completion date of 07/07. ???? ???????? ????????????????????: In preparation for our MVP, we've developed 15,000 simulated customer profiles to enhance the training of our machine learning models. This expansive dataset is instrumental in refining our algorithms, ensuring they perform optimally in real-world scenarios. ??? ???????????????????????? ?????? ????????????????: We are planning to further improve our prototype by incorporating an even broader array of product details. This expansion will allow us to fine-tune our model’s accuracy and extend its predictive capabilities. Moreover, our team is working diligently to automate the dermatologist’s scoring method, enabling us to scale our assessments and deliver personalized product recommendations more efficiently. ?? ?????????????? ??????????: Stay tuned for next week’s update, where we will unveil detailed accuracy assessments of our prototype and provide further insights into the final development phase of our classifier and MVP. Thank you for being part of our journey as we innovate at the intersection of technology and skincare, crafting personalized solutions that are both scientifically grounded and user-centric! #SkincareInnovation #AIinHealthcare #PersonalizedSkincare #MachineLearning #DigitalHealth #TechForGood #StartupJourney #Dermatology #DataScience #ProductDevelopment

? ?????????????????? ???????????? ???????????? ? ?? ???? ?????????? ?????? ???? ????????????: ???????? ?????????????????? ???? ?????? As we approach pivotal milestones in our journey, DermaSkan remains steadfast in delivering our innovative skincare solutions, scheduled for a customer-ready MVP by 7/22! ?? ?????????????? ????????????: We're excited to announce that our prototype development is on schedule, with the prototype set to be completed by 7/4. Our team has also crafted a machine learning model to classify moisturizers based on their ingredients and suitability for specific skin concerns, with validation by a dermatologist set. ???? ???????????????????? ???? ????????: Leveraging a robust list of ingredients identified by our team, the classifier model doesn’t just take claims at face value but provides a scientific assessment of product efficacy. This model will soon be enhanced with actual dermatologist validation data from 150 moisturizers. ?? ?????????????? ?????????? ??????????????: Our scoring model, crucial for understanding product fit for individual customers, is expected to be completed by 7/4. It uses a nuanced 4-point system to rate suitability, from "not recommended" to "excellent." Initial tests using 25 products have already been conducted, with further validation by our dermatologist on an expanded dataset. ??? ???????????? ?????? ???????? ??????????????????: This week, we are excited to welcome new talent to our front-end development team, reinforcing our capabilities to bring you an intuitive and user-friendly digital experience. ?? As we refine and enhance our models, stay tuned for next week’s update, where we will share detailed accuracy assessments and insights from the dermatologist validation process. Thank you for being a part of our journey towards revolutionizing personalized skincare with AI-driven insights! #DermaSkan #SkincareInnovation #SyntheticData #MachineLearning #AI #RecommendationSystem #MVP

???????????????????? ???????????? ?????????????? ??????????? ?????????????????? ???? ??????: ???????????????????????? ???????? ?????????????????? ???????? This week, we're excited to share major strides in our journey from prototype to a minimum viable product (MVP), scheduled for customer-ready release by 7/22! ????????????????? ??????????????: We've detailed the steps necessary to evolve from the initial prototype, set to be completed by 7/3, to an advanced MVP. Our MVP will not only be equipped with a front-end UI but will also harness the power of over 10,000 rows of data to offer alternative product recommendations when scores are low. ??????????????????????? ????????????????: To streamline our development, we've adopted a rule-based method that emulates the decision-making of dermatologists. This approach will generate synthetic data to train our models efficiently. Moreover, we commit to rigorous spot-checking and manual validation to ensure the authenticity and reliability of this synthetically produced data. ??????????????????? ?????? ??????: Our team is implementing a sophisticated algorithm, informed by dermatologist insights, to introduce a "Skin Concern" column in our product dataset. This algorithm intelligently categorizes over 641 moisturizers based on skin type (Dry, Oily, Normal, Combination) and links them to corresponding concerns like aging, dryness, and acne. We will manually verify the algorithm's accuracy by examining 100 random rows and incorporating dermatologist feedback. ????????????????? ?????? ????????????????????: Our development is centered around a robust scoring system to assess product matches. Post preliminary internal tests, a selected data subset will undergo dermatologist review to fine-tune our AI model's accuracy. ?????????????????-?????????????? ??????????: In the coming week, our focus will be to finalize the "score" column in our synthetic dataset for the prototype with dermatologist endorsement, propelling us towards training our AI model. We will also continue to expand our synthetic customer profiles to enrich our MVP data. ???Stay tuned?as we push forward in our mission to transform skincare through AI, making personalized product recommendations not just a possibility but a reality. #DermaSkan #SkincareInnovation #SyntheticData #MachineLearning #AI #Personalization #TechInSkincare