Shelfer is an AI-powered Retail OS for underserved SMB retailers to reduce manual operations by 80%, automate replenishment, and boost sales to neighborhood.- Landed 70+ retail users, securing 38 trial agreements within the first month.- Fully launched on iOS and Play Store, with London-based retail chains piloting the platform.

🔗yt.seanchen.ioI make videos about building tech products and startups. I enjoy working on 0-1 problems and take 120% responsibility in what I believe in building. I also share interesting conversations I had with friends and great connections in my network.

Product Data Scientist at YouTube Shopping, focusing on enabling YouTube VODs, Shorts and Live to become 'shoppable' destinations.

Antler UK 12th Cohort

Building Generative AI & Optimization products for retail & merchandising

We pitched to YC and didn't get in. But it was a great start!:)Vision:- Our tool is a great companion for language teachers so that teachers can customize education by tailoring their language lessons to enhance teacher-student interaction.MVP:- Assist language teachers with creating personalized content based on students' interests and their commonly made mistakes.- Help teachers summarize and organize class materials (recordings) and students assignments- Integrate school's curriculum with our AI tool to free the teachers from the cognitive burdens of remembering the details of the requirements

Product Data Scientist at YouTube Shopping, focusing on enabling YouTube VODs, Shorts and Live to become 'shoppable' destinations.

Product Area: Google Business Communication (BizComms), Google Workspace, re-org to Android- Responsible for data analytics for two pillars: Merchant Success and Monetization; Functional as Data Product Manager for User-Facing Product Areas- Collaborated extensively with PMs, BDs and Eng from Google Maps, Google Search Ads, Google Shopping and Android Messages- Leveraged Data Science to drive BizComms merchant growth, improve onboarding experience and explore new cross-app entry point and monetization opportunities (Conversational Commerce, Click-to-Message Ads, Unified Inbox)

DuoPath.ai is a data science collaboration platform that helps data practitioners and domain experts work together smoothly and seamlessly, leading to more productive business/scientific understanding and outcomes.

Co-founder and co-host of Startup Connect SeriesCo-chair of 2021 Global Chinese Entrepreneur Association (GCEA) Online Competition, hosting top VCs and startups from the globe.

Generated patient-specific embedding using NLP and representation learning techniques using PyTorch for top 50 diseases in the US, which were utilised for predicting undiagnosed diseases via transfer learningConducted a Next Best Action (NBA) project in early prediction of Lower Back Pain surgeries and early intervention with professional guidance and care for our patients (Leveraging big data tools such as Spark/Dask/HiveQL on HDFS)

[Digital Therapeutics — Wearable Enabled Music Therapy Platform]Muser is a wearable-enabled music therapy platform that combines music-streaming service with biometric data analytics to assist music therapists’ treatments and smoothen patients’ healing process. Among the 40 million adults or 18% of the total population in the USA who suffer from anxiety, only 37% receive treatment. Muser leverages the biometric wearable sensing and cloud analytics to quantify, manage and improve the real-time effectiveness of music therapy to people who suffer from anxiety.MIT Sandbox Innovation 2020 Fellow + MIT DesignX 2020 Finalist + Berklee Music and Health Innovation Challenge Finalist (pitched during Boston HubWeek in October. 2019) + 2019-20 IDEAS Social Innovation Challenge Finalis

- Optimizing the assortment strategies for revenue maximization across 12000+ US retailers serving from AB-InBev’s portfolio of ~400 products - Designing and implementing predictive models (Cross Nested Logit) and robust optimization algorithms (Franke-Wolfe) on Azure Databricks using PySpark, PyTorch to increase AB-InBev’s revenue by 5 million dollars

1. Help to identify and network with leading speakers in the Data Science and Operations Research industry for the MBAn Analytics Speaker Series (held monthly during the Fall and Spring semesters)2. Organize logistics with industry speakers – be key point of contact the day of each event3. Promotes the Analytics Speaker Series to fellow classmates across MIT4. Identify additional engagement opportunities for the MBAn class with corporate partners5. Follow-up with speakers with attendance sheet after each event and gather feedback for employers

Machine Learning for Web Traffic Time Series Forecasting with Azure environment support

Build a Dirichlet Process Infinite Gaussian mixture model using Edward and PyMC3 for clustering 12 million customer expenditure data, embedding time-series autoregressive model into stick breaking process for dynamic parameterisation.Use Python with Spark for distributed computation

Our team from UCL ChangeMakers project continued to work over the summer as research assistants at The Alan Turing Institute:- Build an R6 based unified interface (distr6) for probabilistic distributions in an Object-Oriented manner (https://github.com/alan-turing-institute/distr6) - Extend its application to set up probabilistic machine learning pipelines in the future

[Program: UCLChangeMaker - Department of Statistical Science]Our team has previously developed an interactive online platform using ShinyDashboard to allow data exploration, automate statistical and machine learning modelling, and self-learning of key data analysis concepts. Part of our team have built an R Shiny based web app for login system and smart space allocation at UCL Department of Statistical Science.Here are our web apps and Github link:https://egon.stats.ucl.ac.uk/users/zcaksc0/UCLShinyStats/https://github.com/ShenSeanChen/UCLChangeMakers-StatisticalVisualisation-RShiny

- Automate routine data mining (via R & VBA) of Rsam datasets concerning application vulnerabilities detected by penetration testing- Generated weekly burndown reports on 9 transaction cycles regarding the impacted applications

During the summer of 2017, we spent 10 weeks working with Education Partnerships Africa as volunteers in Kenya in order to improve the quality of education for the local young people. Our group worked in direct partnership with disadvantaged rural secondary/high schools. We invested directly in resources aimed at improving the quality of education and welfare received by the students: 1. A new library was added to Riyabu Secondary School in Magena, Kenya: having contacted the local carpenters & welders, managed to make two 12 * 6 * 4 - feet shelves where the newly added, catalogued books were stored; for safety reasons, wire mesh doors with pad locks were installed on the shelves and a new steel door was made for the library; the students would be able to borrow textbooks from the library with an established lending system, finish their assignments in the public study environment, and enjoy reading literature. 2. We also added metallic beds (originally wood ones) to the girls’ dorm where on average two people were sharing one bed for boarding. 3. We held Q & A sessions with the students from all years, mainly talking about studying, career and life at school, through which we gained better insight into what the students really needed

• Interpreting (Chinese-English) for Deputy Premier & Minister for Education(State Government of Victoria, Australia), Minister Counsellor(Education and Research, Australian Embassy, Beijing), CEO of Victorian Curriculum and Assessment Authority during media interviews and tea breaks• Presented as a VCE student representative, sharing experience of VCE in China/Australia• Assisting with arranging the sites of conferences, dinners and tours

• Helping VCE(Victorian Certificate of Education) students with Mathematical Methods(CAS), Further Mathematics, Specialist Mathematics, Physics and Chemistry; British and Australian university application consultation.