Most software was written for people to read. We are rebuilding a large, mature codebase so that AI agents can read it, change it and ship it too, without breaking what already works.

In this role you will decide how the system should be cut up and what each piece promises to the others. You will put legacy logic behind clear contracts, one safe step at a time. You will design the data models, APIs and tests that keep behavior correct under load, retries and failure. You will build the MCP tools and agent context that agents rely on, and you will set the bar for what AI-written code must meet before it merges.

You will work hands-on, not from the sidelines, and you will own your work from design through production. You will also explain your trade-offs clearly to engineers, QA, operations and product, and bring them along with your decisions. The role is remote and on contract, with a globally distributed team. It suits an architect who reasons from first principles, uses AI tools daily with healthy skepticism, and is comfortable with open-ended problems.

WHAT YOU WILL DO

Design systems that agents can operate

- Break legacy monoliths into modular systems with strict boundaries, single responsibility and clear contracts.

- Lead incremental modernization using the Strangler Pattern, wrapping legacy logic behind modern, contract-first interfaces.

- Design data models, keys and constraints from the queries and rules the system must always uphold, and enforce integrity in the database instead of leaving it only to application code.

- Apply SOLID principles, dependency injection and hexagonal architecture so AI-generated changes stay deterministic and regressions stay low.

- Design for real-world cases:



simultaneous requests, retries and duplicate messages, partial failures, and changes to data that already exists.

Build the agent and MCP layer

- Build and maintain the agent context layer: shared skills, rules and commands for AI-driven engineering workflows.

- Build and run Model Context Protocol (MCP) servers that expose legacy APIs, services and databases as typed, permissioned, auditable tools.

- Own contract-first API design as the main interface between human intent and agent execution.

- Design agent loops with clear tools, permissions and stop conditions, and tune them for cost, speed and quality.

Keep AI-written code trustworthy

- Review AI-generated pull requests for correctness, security, scalability and maintainability. Check that tests assert real behavior and do not just pass.

- Require test-first development and characterization tests that preserve legacy behavior during refactoring.

- Set the review habits, checklists and guardrails that help the wider team use AI tools well, and know when not to use them.

Own outcomes and work across the business

- Take responsibility from design through production, including monitoring, incidents and follow-up.

- Ask the right questions up front, explain trade-offs in plain language, and bring product, QA and operations teams along with your decisions.





- Mentor engineers and set a consistent standard for design and code quality.

WHAT WE ARE LOOKING FOR

- 6+ years in software architecture, platform engineering or technical leadership.

- Proven experience modernizing large, poorly documented legacy systems.

- Strong hands-on depth in TypeScript and Node.js, or in .NET. Depth in one is enough.

- Strong command of API design, relational data modeling, SQL, transactions and concurrency.

- Solid system design fundamentals: queues, caches, background workers, sync versus async, and what breaks first as load grows.

- Practical, day-to-day use of AI coding tools on real projects, with a deliberate workflow, a habit of verifying output and a clear view of the risks. Experience building agents, MCP servers or AI tooling is a strong plus.

- A bias toward clean code, tests first, deterministic systems and production-grade AI.

- Transparent written and spoken communication, and the independence to work with little day-to-day oversight.

NICE TO HAVE

- Experience with remote-first or globally distributed teams.

- Background in SaaS transformations, scale-ups or private equity-backed environments.

- Comfort in high-ambiguity, high-ownership settings.

HOW WE HIRE

Expect an online assessment followed by technical conversations where you will reason through data models, queries and designs in plain language, and explain how you use AI tools in your own work. We are looking for sound thinking and honest answers, so please be ready to explain your own resume claims in detail.

Skills:- TypeScript, .NET, NodeJS (Node.js), Generative AI (GenAI) and Agentic AI

📌 Ai Systems Architect (India)
🏢 FAiHr
📍 India