AI Developer in Rajshahi, Bangladesh

My work centres on production AI: software that uses large language models to actually do something for a business rather than demo well. That includes AI agents that call tools and complete multi-step tasks, retrieval systems (RAG) that answer questions from your own documents, chat and voice assistants inside web and mobile apps, and automations that take repetitive work off a team. Around that core I build the product itself — the Next.js front end, the NestJS API, the PostgreSQL database, Stripe billing, and the deployment.

For local businesses the request is often simpler: a booking system, an inventory or customer database, a website that ranks, or a mobile app for customers. I built EasySearch, a local services marketplace for Rajshahi, on exactly this stack, and I am happy to start with the practical problem and add AI only where it earns its place.

Most software work in Rajshahi still goes to agencies in Dhaka or to whoever is cheapest online. Working with a developer in your own city changes the dynamic: we can meet at your office or a café, I can see how your team actually works, and there is no time lost to translation between a salesperson and the person writing the code. I do discovery meetings in person anywhere in Rajshahi and then run the build the same way I run remote projects — a shared board, weekly written updates and a demo at every milestone.

I also mentor and occasionally hire from the Rajshahi University and RUET computer science community, so if a project grows beyond one person I can bring in a second pair of hands I have already vetted, rather than a stranger from a marketplace.

The majority of my revenue comes from clients outside Bangladesh, and Rajshahi is a perfectly good place to serve them from. My working day in UTC+6 overlaps the full UK and EU afternoon and both ends of the US day; I keep a fixed call window for each client and handle the rest asynchronously. Fibre internet, a proper home office and a backup connection mean calls and demos are reliable.

Remote clients get the same terms as anyone else: a fixed-price proposal with milestones, a contract with IP assignment, invoices in USD payable by Wise, Payoneer or bank transfer, and code that lives in their own GitHub organisation from day one. Several of my case studies — Virtual Client, Contently AI, MedScribe — were built for clients I have never met in person.

I publish fixed prices so nobody has to guess. AI integration into an existing product starts at $799 for a single feature delivered in one to two weeks. An MVP starts at $999 for one core workflow and goes to $2,999 for a launch-ready product with payments and an AI feature. AI agent SaaS products start at $1,499, and the multi-tenant Professional tier with RAG and agent workflows is $4,999 over five to eight weeks. A marketing website or landing page built to rank is $499.

For anything that does not fit a package — a local business system, a longer engagement, or a team that needs part-time senior help — I bill hourly at $100-150 or on a monthly retainer from $500. Rajshahi clients paying in taka get the same scope at the prevailing rate.

Book a free 30-minute call or message me on WhatsApp. If you are in Rajshahi we can make that first conversation a coffee instead. I will ask what problem you are solving, who the users are, and what budget and deadline you are working with, and tell you honestly whether one of my packages fits. Within two working days you receive a written proposal. Work starts the week after you approve it.