Frank Villavicencio@fvillavicencio
1
Events attended
1
Submissions made
United States
20+ years of experience
About me
Helping fast‑growing tech companies turn complex ideas into high‑impact, scalable products—with the flexibility of a fractional consulting model. As a Fractional Chief Product Officer, I partner with founders, CEOs, and product leaders to remove blockers, align your teams around a shared vision, and rapidly deliver the product outcomes that boost revenue, increase retention, and prove market fit. Whether you’re: • A FinTech or HRTech founder juggling compliance, integrations, and investor expectations • A Cybersecurity innovator struggling to translate technical breakthroughs into enterprise-ready products • An AI-first startup moving from research to real-world use cases • Or a health-tech or research org trying to build user-friendly, regulatory-ready tools— I provide seasoned product leadership, fast. With 20+ years of experience, awarded patents, and a background across regulated and fast-moving industries, I partner with my clients to: • Launch products faster with less risk • Align stakeholders around a clear product vision • Translate deep tech into intuitive user experiences • Build scalable systems without losing agility • Deliver investor confidence and customer trust My fractional model gives you immediate, cost-effective access to the leadership and execution horsepower you need, tailored to your stage, stack, and strategy. Specialties: • Fractional Product Leadership & Strategic Advisory • SaaS Product Engineering & Lifecycle Management • FinTech • HRTech • Innovation and Execution • Identity & Access Management (IAM) Strategy • AI and Digital Transformation
🤝 Top Collaborators
.png&w=640&q=75)
Tashi Yang
🤓 Latest Submissions
ProductAdvisors: Agent that cuts what it measures
ProductAdvisors is a four-sleeve autonomous trading agent on one Alpaca paper account (PA3TPECBLC29, $100,000 from 28 August 2026). Every entry, exit, size and stop is deterministic math. An LLM never places an order. SIXFOLD (75%) buys quality names from a six-lens fundamental score. A three-model Advisory Council on our self-hosted Dell4 cluster then votes in parallel: dell4-finance (Fin-R1 7B), dell4-chat (Qwen3.6-35B MoE), dell4-qwen38 (Qwen3.8-27B). Two of three must APPROVE. They can only veto; they cannot pick a ticker, set size, or submit to Alpaca. If fewer than two respond, the quantitative signal proceeds. Exits never wait for a vote. CSP (15%) sells cash-secured puts. Quotes come from the official Alpaca MCP server; a contract it cannot quote is never priced. Pendulum (5%) is TLT mean reversion using the same decide() live and in backtest. Vampire (5%) is a QQQ/TQQQ micro-scalper: retired on day 5 after the broker ledger showed −$663 over 10,470 fills while the engine claimed a profit, then re-admitted on day 7 behind an LLM regime gate. dell4-chat labels the last 30 one-minute bars every 15 minutes; new lots open only in chop. A missing, stale, or unparseable verdict closes entries. Exits never wait. A narrator (dell4-chat) writes session prose after the fact; a failed narration costs a paragraph, not a trade. Risk is independent of the models: 5% per trade, 10% per position, $5,000 cash floor, a 2% daily-loss breaker that a restart cannot re-arm, fills polled to a terminal state, and a watchdog that reads the broker every 20 seconds. As of 3 September 2026, 15:25 ET the paper book was $100,552 (+$552). A 30-day bars-only replay of this allocation reached $101,085. We do not claim the dashed CSP extrapolation as official P&L. Judges should score the paper account. Live dashboard: http://44.202.110.127:8501/ Repo: https://github.com/mihai-satmarean/alpaca-trading-agents Team: Frank Villavicencio, Mihai Satmarean, Tashi, Vibhuti Singh
4 Sep 2026
