Empowering Innovation: Highlights from the Women in AI RAG Hackathon
Retrieval-augmented generation (RAG) is transforming AI applications by improving the accuracy and reliability of generated responses. Unlike traditional AI models that rely solely on pre-trained knowledge, RAG enhances outputs by dynamically retrieving relevant information from external sources. As AI adoption accelerates, mastering this technology is key to building smarter, more context-aware applications.
On January 25, 2025, the inaugural Women in AI RAG Hackathon brought together a diverse group of women technologists at Stanford University to tackle this challenge head-on. Organized by Zilliz, GenAI Collective, and Women Who Do Data (W2D2), the event provided a hands-on opportunity for participants to experiment with cutting-edge AI tools and collaborate with fellow innovators. Over the course of the day, teams built working RAG-powered applications using the Milvus vector database—many of them solving real-world problems in healthcare, legal access, sustainability, and more—all within just a few hours.
While most of the time was devoted to heads-down hacking, the event featured talks from veteran technologists who shared insightful advice and encouragement during the lunch session, energizing participants for the competition and underscoring the importance of community and collaboration.
Technical mentors worked with the teams throughout the day to answer questions, debug code and provide encouragement. The judging panel of business and technical leaders from the AI industry provided constructive feedback and ideas for the continued development of the applications.
Thank You Sponsors
This event was made possible by the support of our sponsors— AWS, TwelveLabs, Arize, OmniStack, StreamNative, and Mistral AI—who contributed prizes to recognize the outstanding projects, including:
$10,000 in AWS Credits
$2,200 in cash
$500 in Mistral Credits
$5,000 in OmniStack inference credits
$2,000 in StreamNative Cloud credits
Built with Milvus
In just a few hours the hackathon teams were able to create working prototypes of apps that spanned personal wellness, democratizing legal access, increasing sustainable building to helping inspire creativity with legos and social media marketing.
App: Skinnify (1st place)
Team: Sandhya Sangli, Shirley Luo, Stuti Kafle, Sanjana Gajendran
It can be hard to get a dermatologist appointment - it can even take months! Skinnify is a RAG-based model for personalized skincare product recommendations. It allows you to upload a photo of your skin and get insights into your skin. It even gives advice on products to use to help with skin problems like acne.
App: CourtIQ (2nd place)
Team: Prachi Sethi, Oshleen Gupta, Daniella Pontes
CourtIQ addresses the “Justice Gap” by helping individuals in situations that could qualify as a legal case to quickly assess their cases and identify the most skilled lawyers with the help of an AI agent overcoming the usual struggle to take action. The app provides quick case profiling and categorization, a list of lawyers focusing on the case domain, and case summarization with key supporting elements.
App: cycle (3rd place)
Team: Jessica Singh, Jasmeet Bajwa, Subhiksha Mani
cycle addresses the challenge of inaccessibility and lack of personalization in wellness plans for women. It's tailored to the four phases of the menstrual cycle (menstrual, follicular, ovulatory, luteal). Cycle is meant to provide women with specialized workout and diet plans that are synced to their cycle and profile to better suit the the monthly variations of energy levels and dietary needs.
App: Brickspiration (Best use of Mistral)
Team: Deepika Khammampati, Nidhi Pai, Meera Tresa Sebastian
Create endless LEGO creations with a chatbot that helps repurpose LEGO sets for different builds. Brickspiration suggests creative and compatible alternative builds from existing LEGO sets. It uses semantic search to output similar pieces.
App: Compliagent
Team: Akhila Josyula, Roxana Raicu, Meghna Natraj, and Meghna Pusala
Compliagent is the backbone of a trustworthy and efficient organization, protecting against risks, penalties, and reputational damage. The challenges with compliance are small and expensive compliance teams, multiple stakeholders, high workloads, lagging response times, and delays in compliant product releases. The app, Compliagent, becomes part of the compliance team (human-in-the-loop). It answers real-time queries from customers regarding company compliance policies and more. The agent can be added to Slack to answer employee questions.
App: RSRCH
Team: Atisha Rajpurohit, Ananya Gupta, Melody Masis, Ashley Rice
There are a lot of research papers out there (2.4 million!) and it’s hard to read and understand all of them. RSRCH took 100,000+ computer science articles and turned them into a chatbot to understand them better by asking questions such as “I am a [ __ ], what is the most up-to-date research being done in [ __ ]”
App: Workplace Detox
Team: Mrunmayee Rane, Sophia Giglotti, Sri Harshitha Avasarala, Supriya Ramarao Prasanna
When you’re stressed at work and have nobody to talk to, you can ask for advice from this chatbot! Workplace Detox is a chatbot that answers your stressful work situations and gives you advice from books, videos, and blog articles.
App: CodeSolve
Team: Emily Moini, Emiko Sano, Neha Sharma, Svea Meyer
Developers spend hundreds of hours creating and managing issues and bug tickets. A GitHub repo is a historical record of your project. Going through it page by page can be time-consuming. ColdSolve is one interface with seamless knowledge management across your GitHub repo. You can search for past issues and find similar solutions.
App: RosieRAG
Team: Jennifer Tran, Silvana Smoiceanu, Iris Yu, Lijie Zhou
Small businesses have limited marketing budgets, but 41% of them rely on social media marketing to generate sales and foot traffic. RosieRAG analyzes all influencer-made videos produced by clients and extracts the most engaging and relevant segments. The final output is a short, impactful reel or video — approximately 20 seconds long — designed to capture attention instantly.
App: Professional Matchmaker
Team: Sharvari Deshpande, Junie Varghese, Ji Young Lee
Professional Matchmaker is a job matching system that leverages Milvus, an open-source vector database, and LangChain, a powerful library for creating LLM-powered pipelines, to match job postings with candidate resumes. By utilizing Sentence-Transformers for embedding generation and Milvus vector search, it retrieves the most relevant jobs from a dataset of job postings and generates an explanation for why each job is a good fit for the candidate.
App: Greencode Labs
Team: Liana Soima, Lourdes Lopez, Leigh Miller
The path to sustainable construction is riddled with challenges: intricate codes, limited resources, and missed opportunities for eco-friendly improvement. Builders need a way to simplify this process and set themselves apart as leaders in sustainability. Greencode Labs exhibits regulatory compliance for eco-friendly buildings, powered by AI. The app validates environmental compliance effortlessly and provides actionable recommendations to optimize sustainability. With Greencode Labs, builders don’t just build homes - they set a standard for the future.
What’s Next for the AI Women in RAG Hackathon?
The success of this event highlights the growing need for inclusive spaces that provide hands-on experience with emerging AI technologies. We’re excited to see how these apps continue to develop and startups that grow out of these efforts.
We look forward to hosting future hackathons and continuing to support innovation in AI. If you missed out this time, keep your eyes open for virtual trainings and other upcoming Unstructured Data Meetups and GenAI Collective and W2D2 events.
Emily KurzeEmily Kurze is the Director of Marketing at Zilliz.
Start Free, Scale Easily
Try the fully-managed vector database built for your GenAI applications.
Try Zilliz Cloud for FreeKeep Reading
How to Choose A Vector Database: Weaviate Cloud vs. Zilliz Cloud
Compare Weaviate Cloud vs. Zilliz Cloud and Milvus in this in-depth benchmark, cost, and features comparison.
Contributing to Open Source Milvus: A Beginner’s Guide
Contributing to open source software is a rewarding way to improve your programming skills, collaborate with others, and give back to the development community. Learn how to contribute to Milvus with this beginner guide!
Advanced Video Search: Leveraging Twelve Labs and Milvus for Semantic Retrieval
In August 2024, Twelve Labs and Milvus (vector database by Zilliz) joined hands to create powerful video search applications.