AI Product Engineer & Technical Founder — I build AI-powered products from zero to production.
15+ shipped products. 7,100+ GitHub contributions in the last 12 months.
Public projects
m0lz.00 — Monochrome developer blog & portfolio. Static-only Next.js 16 site with MDX content pipeline, generative branch mark identity system, and dark/light theme. Next.js / TypeScript
m0lz.01 — Automated content publishing agent. Commits MDX posts to m0lz.00, cross-posts to Medium and Dev.to. TypeScript
m0lz.02 — Structured AI coding workflow orchestrator. Plan, Implement, Contract-Evaluate with dual-model adversarial evaluation, WISC context management, and tiered verification. Rust / TypeScript
m0lz.03 — Security proxy daemon for MCP servers. Authentication, rate limiting, PII detection, permission scoping, audit logging. 97% detection rate across 7,095 attack scenarios. TypeScript
m0lz.04 — AI-powered legal case management for pro se litigants. Court filing monitor, deadline tracker, citation verification, red team analysis, and DOCX export. Local-first, privacy-focused. JavaScript / Claude API
Investor Matchmaker — Investor-founder meeting scheduler for Startup Week. Greedy bipartite matching across investment type, business type, and sector with Excel I/O. Python
Private projects
Bloom — AI-powered revenue discovery for service businesses. Satellite measurement, AI proposals, Bloom Intelligence email personalization. TypeScript/Next.js
Alpaka — Value chain intelligence for real estate. Scope 3 carbon emissions ML modeling, supply chain optimization, vendor management. Python ML + TypeScript
Ready Text — Waitlist texting & customer alert platform for restaurants and businesses. Live SaaS with paying customers. Laravel/PHP
How I build: m0lz.02
m0lz.02 implements Plan, Implement, Contract-Evaluate — a structured methodology for AI-assisted engineering that uses:
- Contracts between human and AI defining "done" before any code is generated
- Dual-model adversarial evaluation — Claude evaluates contract compliance, GPT challenges the design approach. Different model families = different blind spots.
- WISC context management — Write, Isolate, Select, Compress — for managing the most precious resource in AI-assisted development
- Tiered evaluation rigor scaled to change risk
Read the framework: github.com/jmolz/m0lz.02
Tech stack
Languages & Frameworks: TypeScript Python Next.js FastAPI Laravel Django React
AI/ML: Claude API OpenAI API Multi-Agent Architectures MCP Servers LLM Orchestration RAG Pipelines Embedding Models Rerankers scikit-learn Computer Vision
Model Deployment & Infrastructure: Self-Hosted Open-Weight Models Quantized Models for High-Throughput ETL Full-Weight Models for Deep Analysis RunPod GPU Docker-Based Model Serving Secure Closed-Environment Inference
Infrastructure: Docker PostgreSQL Supabase Redis CI/CD
The numbers
- 7,100+ contributions in the last 12 months
- 15+ products built from zero to production
- 4 companies co-founded with technical leadership
- 2,900+ commits on Alpaka alone
Previously
15 years of B2B SaaS leadership (VP/SVP/CSO) at companies including 24/7 Software and CleanAir.ai — which means I don't just build AI systems, I understand why they matter and how they create business value.
MBA — Nova Southeastern University
BSBA, Entrepreneurship — University of Central Florida