Cut customer-support headcount 70% with a tier-1 agent
Problem. Support team of 12 was drowning in repetitive Tier-1 tickets. Founders wanted to scale users 3× without growing the team.
Senior full-stack & AI engineer. I help founders and engineering teams ship scalable AI agents, RAG systems, and full-stack apps that actually survive contact with real users.
Multi-step, tool-using agents that actually replace repetitive work — built with LangChain, CrewAI, or LlamaIndex on top of OpenAI / Claude. Not chatbots. Workers.
End-to-end web apps with AI baked in — auth, billing, dashboards, vector search, real-time UX. Next.js or MERN on the front, FastAPI or Node on the back, Postgres or Mongo behind it.
Retrieval-augmented systems that ground LLMs in your data — chunking strategy, embeddings, hybrid search, eval harnesses. The part that decides whether your AI feature is useful or just confident.
I audit your workflow, find the repetitive 60-80%, and replace it with agents and automations. Concrete deliverable: a written plan with hours saved and dollars cut, then I build it.
Problem. Support team of 12 was drowning in repetitive Tier-1 tickets. Founders wanted to scale users 3× without growing the team.
Problem. Compliance team needed to answer policy questions against 4M+ regulatory PDFs without leaking data to public LLMs.
Problem. Patient onboarding took 47 minutes of clinical staff time per case. Repetitive form-filling, EHR lookup, eligibility checks.
Problem. Legacy PHP storefront was buckling at 50K MAU. Needed a rebuild — and the founders wanted AI-driven product discovery, not just a port.
† Client names redacted under NDA. Reference calls available on request after a discovery call.
We figure out if I can actually help. If I can't, I'll point you to someone who can. No deck, no sales pitch.
Written proposal: problem, approach, stack, milestones, fixed price or T&M. You sign or you don't.
Daily commits, weekly demos, shared Linear board. You see progress every day. No black boxes, no surprises.
Production deploy, observability, runbooks, handoff session with your team. 30 days of free bug-fix support after.
I'm a senior full-stack & AI engineer with 5–7 years of production experience. I started as a MERN dev, went deep on Python and ML, and now spend most of my time at the intersection — turning AI prototypes that barely work in a notebook into systems that survive real users, real load, and real money.
I work best with founders who want one senior generalist instead of three juniors, and with engineering managers who need a surgical contractor — not someone they have to onboard for a month.
I won't take a project I don't think will succeed. If your idea needs a different shape, or doesn't need AI at all, I'll tell you on the discovery call. Wasted time is more expensive than an honest no.
For a defined feature, MVP, or migration with clear boundaries.
Ongoing senior bandwidth — your roadmap, my keyboard.
I review your stack, workflow, and AI surface area. You get a written plan.
† Prices are starting points. Larger or more complex engagements are quoted after a discovery call.
No deck, no script. We'll figure out if I'm the right fit. If I'm not, I'll send you two people who probably are.