AI integrated into the product you already run
I add AI to existing software: OpenAI and Anthropic APIs, RAG over your own data, CRM and API integrations and MCP servers, wired into your product without breaking what already works.
Who this is for
- SaaS teams adding AI features to a product that already has users
- Businesses that want AI working on their own documents, records and tools
- Teams with a demo that works once and breaks on real data
What I do
- OpenAI and Anthropic integrations with structured output your code can trust
- AI features inside an existing SaaS: same auth, same tenants, same data rules
- CRM and third-party API integrations behind swappable adapters
- RAG and knowledge-base systems over your own documents and data
- MCP servers that expose your systems to AI tools safely
Proof: products I shipped
- PavlySaaS · founder build
Small shops sell through reels and DMs, then track orders by hand.
- PhloSaaS · pharmacy
Indian pharmacies key in every purchase bill by hand and juggle GST, expiry and stock in separate tools.
- UxlyzeSaaS · AI analysis
Teams guess why landing pages do not convert.
How an engagement runs
- 01
Discovery. I read the code, the data and the bill, and talk to the people with the problem. You get a short plan: what is broken, what ships first, what waits.
- 02
Engineering. Working software on your real data in days. Fix what is broken, build what is missing, replace what will not hold.
- 03
Integration. Wire it into what already runs: APIs, databases, auth, payments, CRMs, MCP servers.
- 04
Deployment. Your cloud, your accounts. CI/CD, environments, monitoring and backups, so launch day is not the first deploy.
- 05
Support. Docs, handover and iteration. I own the outcome after launch, not just the merge.
Related writing
- How to Take an AI-Built App from Prototype to Production
A working checklist for apps built with AI coding tools: secrets, data access, config, payments, background work, deploys and monitoring.
- Why AI Agents Fail in Production and How to Fix Them
Eight ways AI agents break once real users and real data arrive, how to recognize each one, and the fix that holds.
- How Uxlyze audits a landing page with headless Chrome and Gemini
The architecture behind an AI UX audit tool: a Go worker, chromedp, a job queue and structured LLM output.
- Build the SaaS before you have the API keys
How Pavly runs every third-party integration behind adapters that default to mocks, and why keyword rules run before the AI.
Common questions
- Will our data be sent to a model provider?
- Only what the feature needs, and you decide which provider. Where data cannot leave your infrastructure, I can deploy an open-source model you host yourself.
- How do you keep model costs under control?
- Cheap rules first, the model as the fallback. In Pavly, keyword rules classify most comments and the model only sees what the rules miss.
- What is an MCP server and do we need one?
- It is a standard way to give AI tools access to your systems with defined tools and permissions. You need one when you want assistants or agents to act on your data instead of reading pasted copies of it.
Other services
Have an AI product to ship?
Tell me what needs to be built, fixed, integrated or deployed. You work with me directly, the engineer who reads the code and ships it.