Previously built at early-stage startups, shaped strategy for Fortune 100 companies, and scaled a business from $0 to $75K in revenue.
Hermes monitors neighborhood feeds across 8+ accounts, scores every post for service intent with an LLM layer, and replies to qualified leads in each account's local voice, tracking every lead from first reply to booked walkthrough.
I spotted the opportunity and helped propose and develop Hermes, ran the system day to day, and worked with engineering to instrument lead latency when posts started surfacing hours late.
Tracking 4 expansion markets and 5 Bay Area regions meant pulling numbers together by hand every time someone asked how we were pacing. The dashboards replace that with one live view the whole team shares.
Lead source performance, MQL to walkthrough movement, pacing against member targets, plan mix, and a pipeline forecast for the month.
Before Oski, every match meant updating 15+ spreadsheets by hand, downloading each one, and re-uploading everything to a shared drive. Getting stats to coaches simply took too long, every single time.
Every serve, pass, and attack gets coded, 200+ interactions a game, into one site. AI generates insights from the data, and coaches scout opponents with the same stats and analysis.