Running an AI Hackathon In San Francisco
I had the privilege of co-organizing an AI-focused Hackathon in San Francisco last weekend, as well as mentoring the teams there and being a judge for the final demos.
Few months ago, two of my friends (Ravi Vadrevu, cofounder of Kriya.ai, and Madhweep Lakshmanan, cofounder of lyte.ai) and I had formed a tech community for solopreneurs & entrepreneurs called Hack*Hustle, or H2 for short. Our first event was a simple networking event, which was a great success, and helped us define a direction for the organization as well as form a vibrant Slack community.
Given the trend of venture capital shifting to later stages, accelerators like YC and 500 startups constantly upping their admission criteria, and the general slowdown in the frequency of hackathons, we felt there was a vacuum at the “incubation” stage that we could help address.
Hence, for our second event, we wanted to do something that culminated from suggestions and feedback from our members, and also fill this vacuum, and decided that an AI-focused hackathon would fit the bill.
We had 25+ attendees, which resulted in 6 teams who presented at the demo day. Most of the attendees were mid-to-senior level AI engineers so the hackathon gravitated towards being fairly technical.
The day started with an AI workshop, where the teams were introduced to the AI MVP, the “Lean AI” startup mentality, how to think through AI business models and ROI, and examples of good and bad AI applications.
Over the two days, some great mentors helped us coach the teams, give them feedback, and provide technical and business insights. Our mentors included Erik Chan, founder of RocketClub, Rahul Pattamatta, founder of Supervised AI, and Abhinav Kuru, founder of JoyUp.
We had compiled a list of hackathon resources for the founders, and provided them guidance on how to think through their problem, validating their solutions, researching the market, building a feature/model/wireframe in 24 hours, and pitching to investors on demo day. The recommended deliverables to complete within the two days and prior to Sunday night’s demo day were:
1) a slide deck,
2) a landing page that was used for customer validation efforts, and,
3) a feature (for technical participants) or wireframes (for non-technical participants).
Our demo day judges were a mix of investors and founders, and included: Manisha Jain & John Fohr (both from SandHill Angels), Daniel Gomez (Fusion Fund), Nikhil Aitharaju (2x founder of Tint and Apparatus), Ravi Vadrevu (2x founder of Meed and KriyaAI) and myself (EIR at SymphonyAI).
The resulting presentations were nothing short of extraordinary. The projects covered a wide range of industries, and were a mix of B2B and B2C solutions (skewed towards B2B).
The winning pitch was made by Charan Thimmisetty and Babji Nallamotu who pitched BrieFin, a tool aimed at individual investors that summarizes financial information and personalizes it to the investor’s interest. The investor can read the daily summary or listen to it as a podcast. The team also had a great demo that synthesized Apple’s recent news into a podcast.
The runner up was Ravi Botla with his project, Archirfai. In mid-sized enterprises, engineering teams typically go through multiple rounds of architecture drafting, and get approvals from a variety of teams such as Enterprise architecture team, Information security, IT operations etc. before they develop a new product or work on a major release. The whole process takes multiple weeks to months and is error prone leading to non-compliance with multiple other factors adding complexity. Ravi’s solution was an AI-based application that reviews your architecture based on the enterprise standards, latest developments in technology arena, and then provides feedback and recommends some best practices.
It was incredible to see so many talented folks — the participants, but also the mentors and judges — take time out of their weekend to hack AI projects and/or support the hackers. It gave us (the H2 team) further validation that there is a huge need for early stage incubation support in the market, and we are excited to continue brainstorming ways to solve that problem!
P.S. And a huge thanks to our sponsor: Google Cloud!
