Full-Stack AI Engineer | AI SaaS, AI Agents, MCP, LLM Integration
I've spent 10+ years building full-stack products, and the last few shipping AI products that real customers pay for. Most AI projects fail not because of the AI, but because of everything around it: the architecture, auth, billing, and deployment. I build the whole thing, from scope to shipped.
- AI SaaS platforms, end to end. Multi-tenant architecture, auth, Stripe billing, admin dashboards, plus the AI layer (chat, voice agents, semantic search)
- RAG systems that give accurate, source-grounded answers over PDFs and knowledge bases. Pinecone, pgvector, Weaviate, hybrid search
- AI agents and automation with LangGraph and CrewAI, plus n8n and Make workflows wired into HubSpot, Salesforce, Stripe, and Slack
- Generative media pipelines at production scale. ComfyUI, Flux, LoRA training, ElevenLabs, LiveKit, RunPod GPU
- Magical Children's Books: AI SaaS that turns a single photo into a personalized illustrated storybook using GPT-4o and Flux. Live in production with paying customers
- DATASIM: AI character generation platform running dual ComfyUI pipelines with automated LoRA training. Processes thousands of images daily
- HireHarry: multi-agent platform that works like a virtual employee for marketing, content, and analytics. Deployed on AWS with CI/CD
- FlowMind: AI automation system built with GPT-4, n8n, and FastAPI for intelligent data scoring and routing
Building an AI SaaS, adding AI to an existing product, or stuck with a half-finished MVP? Reach out, happy to talk it through.
