Matt Gacek

Matt Gacek

About

Detail

Founder @ Stealth | AI/ML Engineering | RAG & Agent Systems Architect | Founder (Acq. by Miri) | Building the Future of AI
San Diego, California, United States

Contact Matt regarding: 
work
Full-time jobs

Timeline


work
Job

Résumé


Jobs verified_user 0% verified
  • Owner.com
    Applied AI Lead
    Owner.com
    May 2026 - Current (5 months)
  • S
    Founder
    Squib
    Mar 2025 - Current (1 year 7 months)
    Founded and shipped a multi-tenant AI platform for agencies - from first line of code to $20K+ revenue in 3 months. Built three production services end-to-end: a Next.js dashboard, a tRPC/Express API layer, and an autonomous agent runtime that deploys AI personas across 9 social channels - generating, reviewing, and executing content and engagement without human intervention. Architected a RAG-based persona engine that ingests transcripts, voice samples, and documents into structured memories with embeddings, then retrieves weighted knowledge at generation time - prioritizing how someone sounds over what they know. Built voice fingerprinting, confidence scoring, humanization scoring, and AI-content detection to keep output quality above the
  • Roam
    Builder
    Roam
    Sep 2025 - Mar 2026 (7 months)
    world models
  • MindPalace
    Senior AI/ML Engineer
    MindPalace
    Jan 2025 - Apr 2025 (4 months)
  • MindPalace
    AI/ML Engineer
    MindPalace
    Sep 2024 - Jan 2025 (5 months)
    Architected a RAG system with multi-agent validation workflow - 15 nodes/batch parallel processing, custom temporal fidelity algorithms, 90%+ entity matching confidence. Built a distributed data ingestion pipeline processing +1.5M heterogeneous documents per user across Gmail, Slack, iMessage, Spotify, WhatsApp, and Chrome - sub-12s response times, 99% uptime. Designed entity-aware search refinement using parallel validation agents and custom embedding strategies with ±1 day temporal precision - boosted retrieval accuracy 52%, cut latency from 55s to 10s. Integrated Groq LLM infrastructure (Llama 3.3 70B, 405B) with a provider abstraction layer and dynamic model routing based on query complexity. Built the full frontend in React/Vite - 25+
  • Miri AI
    AI Software Engineer
    Miri AI
    Nov 2023 - Sep 2024 (11 months)
    Built one of the first tool-selection agent systems before function calling frameworks existed — agents dynamically rendered UI components through custom prompt engineering and agent-UI interaction protocols. Redesigned the agent response pipeline with lightweight specialized agents and contextual workflow optimization - cut end-to-end response time from 80s to 6s. Built multi-agent orchestration for collaborative query processing - context summarization, nutritional analysis, adaptive response generation, and personalized coaching recommendations. Trained agents on 6 personal coach profiles to replicate real coaching styles. Ran 5 QA cycles with vendor teams to validate response accuracy and iteratively refine prompt chains. Integrated FDA
  • F
    Founder
    Jul 2023 - Dec 2023 (6 months)
    Acquired by Miri (health/wellness) within 6 months of launch. Founded and built an AI therapy and personal growth app on React Native and Expo - fine-tuned Falcon 40B for coaching interactions and architected an early RAG implementation using Firebase for context retrieval before vector databases were standard. Scaled to 600+ active users in 3 weeks with 21% retention. Ran 100+ user interviews to drive product decisions and behavioral analytics loops. Led a team of 4 shipping full-stack features - personalized user modeling, goal tracking, adaptive coaching workflows. Launched beta across TestFlight and Google Play, 200+ testers in 2 weeks.
  • Pillar4 Media
    UX Research & Product Intern
    Pillar4 Media
    Jun 2023 - Aug 2023 (3 months)
    Owned the A/B testing pipeline - designed experiments in Adobe XD, built variants in JavaScript, and measured against engagement and revenue KPIs. Shipped the GGR Header redesign from research through deployment - 8% conversion lift, 3,000+ new user interactions in first cycle. Drove a 3% RPV increase through Mini Reviews optimization ($1.1M projected annual impact). Scoped a follow-on mobile UX initiative worth an additional $1.3M in identified upside.
  • Envision Capital Management Group
    Quantitative Developer
    Envision Capital Management Group
    Sep 2022 - May 2023 (9 months)
    - Developed scalping and momentum trading algorithms in C#, turning $500 initial investment into $1,200 in 5 trading days while maintaining consistent 73.68% win rate and 2.39 profit factor. - Implemented sophisticated risk management protocol with dynamic position sizing and automated stop-loss triggers, limiting drawdown to 2% per trade while maximizing profit potential. - Built real-time signal detection algorithms focusing on MACD crossovers and RSI divergence patterns, integrating multi-timeframe analysis for trade execution.
  • D
    Intern
    Defend Edge
    Jun 2020 - Jul 2020 (2 months)
    - Enthusiastically shadowed senior developers, gaining first exposure to professional software development practices and team workflows while trying not to spill coffee. - Learned fundamentals of Python development and email automation, while mostly nodding along during standup meetings.
Education verified_user 0% verified
  • Indiana University Bloomington
    bachelor, Bachelors of Science in Computer Science, Computer Software Engineering
    Indiana University Bloomington
This is a community-created genome.