I build AI features that hold up in production.
Senior Python engineer in Colombo, Sri Lanka. I design and ship RAG chatbots, AI agents and LLM features on FastAPI and Django backends. Most recently: Phoneline.ai, an AI agent that answers business phone calls.
Available for new projects, 40+ hours a week
What I build
I work with startups and SaaS teams who want AI features their customers can actually rely on, on backends that won't need rewriting in six months.
RAG and LLM features
Assistants that answer from your documents, website or database and show their sources. Evaluation, cost limits and error handling are part of the build, not an afterthought.
AI agents and automation
Agents that call your tools and APIs to research, draft and update records, with a person approving the steps that matter.
Python backends
FastAPI, Django, PostgreSQL, Redis and Celery. Tested APIs and data models that stay fast as usage grows, deployed with Docker and CI.
Phoneline.ai: an AI agent for business phone lines
Lead engineer, January 2025 to now. I built the product end to end.
Small businesses miss calls because nobody is free to pick up. Phoneline gives them an AI agent that answers from their own website and any other sources they add, like opening hours, prices and policies, and hands the call to a person when it should.
Callers can't see sources or scroll back, so the agent has to be fast, stay on topic and know when to stop guessing and hand over.
What I built
- AI pipelineA RAG system over each customer's website and other sources, prompt design and response evaluation.
- Voice layerReal-time phone conversations on the OpenAI Realtime API, connected to the RAG system, plus call-transfer logic.
- BackendPython FastAPI services, PostgreSQL, third-party integrations, deployment and infrastructure.
- DashboardReact app for sign-up, agent configuration and call logs.
How it works
- Build the knowledge baseWhen a business signs up, its website and any other sources it provides are turned into a searchable knowledge base.
- Talk in real timeThe caller speaks with the agent through the OpenAI Realtime API, voice in and voice out, so the conversation flows naturally.
- Look up the answerWhen the caller asks something, the agent searches the knowledge base for the relevant passages.
- Answer from the business's contentThe reply is based on what the business has published, not the model's general knowledge.
- Hand off when neededIf the caller needs a person, the call is transferred to the team, and every call is logged in the dashboard.
Experience
Lead engineer, Phoneline.ai
Built the AI phone agent end to end. Read the case study.
Freelance Python engineer
- Backend for a US telecom and VoIP SaaS company
- Real-time dashboards with FastAPI and WebSockets
- Built and launched OnCredit, my own lending platform
Senior software engineer, IFS
Enterprise software at a global ERP company.
Tell me what you're building
Send a short description and I'll reply with a concrete first step, usually within a day.