TrioAmigo: senior full-stack, cloud and AI engineering team TrioAmigo

We’re a pre-formed team of three senior engineers. We build full-stack platforms, cloud-native systems on AWS and Azure, and production AI: RAG, AI agents and LLM-powered workflows. Clients bring us in to build a new product, to add AI to an existing one, or to reinforce their in-house team.

Usually this week. There’s no hiring loop, because we already work as a team with shared conventions and tooling. We start with a shadow sprint and reach full velocity by sprint 2.

Two. A fixed project is milestone-based, with a scoped deliverable and a fixed fee, and there’s no set end date: it can run for a few weeks or continue for years, one milestone after another. Part-time gives you 20–30 hours a week per engineer, on an ongoing basis. You can also start with a fixed project and move to part-time once the partnership is a fit.

Yes, and we’re glad to. There’s no upper limit on how long we work together: a fixed project can continue milestone after milestone, and part-time work is ongoing by design. We’re built for the long run, as one team that stays with your product instead of rotating people in and out.

Fixed projects have a fixed fee agreed up front, tied to clear milestones. Part-time work is billed transparently, by the hour or as a package, with one contract and one invoice flow (or per person, if you prefer). Tell us about your scope on a 30-minute call and we’ll come back with an estimate.

We’re in Belgrade and East Sarajevo, on CET/CEST. That means 2–3 hours of morning overlap with US East and full-day overlap with the UK and EU. We work async-first, with Slack or Teams for daily questions and a weekly demo and planning call.

React, Next.js and TypeScript (and Angular) on the frontend; .NET, Node.js with NestJS and Python with FastAPI on the backend. AWS first, plus Azure. PostgreSQL, SQL Server, MongoDB, Elasticsearch and Redis for data, and Amazon Bedrock, RAG with pgvector, AI agents and MLOps tooling for AI. The full list is in our tech stack.

Yes, it’s a big part of our work. We build permission-safe RAG over your documents, LLM features on Amazon Bedrock (Claude) with cited answers, AI agents and integrations such as GitHub Copilot custom agents and MCP, and speech, translation and document pipelines. Governance is designed in from the start, including EU AI Act requirements and privacy-first analytics.

At every stage of delivery: Claude Code and Copilot for scaffolding and boilerplate, AI-assisted review on every pull request, rapid prototyping, documentation and large refactors. On greenfield features that makes us 2–3× faster. Every change is still reviewed by one of us and runs through CI.

Hiring takes 2–4 months per role plus 4–8 weeks of onboarding, and there’s no guarantee the people will click. Agencies tend to rotate staff. We’ve worked together for more than seven years, so you get three senior perspectives, shared conventions from day one, one contract and no single point of failure.

Yes. The part-time model is made for that: senior reinforcement without the overhead of a full engagement. We plug into your rituals and tools, such as GitHub or GitLab and Jira or Linear, and adapt to your process instead of forcing ours.

It’s a no-commitment intro. You tell us about your project and goals, we ask a few questions and suggest how we’d approach it and which engagement model fits. Your request reaches all three of us, and we confirm a time by e-mail.