AI Automation Engineer (Remote - India)

Build the brain behind agentic AI systems.

About the Role

We build agentic AI systems that take real, messy business workflows and run them end to end, with software doing the operational work and people stepping in only at the decisions that genuinely need a human. As an AI / Orchestration Engineer, you build the brain of those systems: the orchestration layer that holds state, decides the next best action, and routes the right agent to the right job at the right time. A lot of the patterns here are still being worked out, so you'll have real room to shape how they come together.

You'll work inside a Monks delivery team, alongside our clients, product, and the domain experts whose judgment the system has to respect. You own the hard part of the product and are trusted to make it real.

We're all-in on agentic engineering. Tools like Claude Code and Codex are part of our everyday workflow, and we're looking for people who are genuinely fluent working alongside AI coding agents and excited to push those practices further.

Who You Are

• A self-starter in the grey areas: you see an undefined problem, design a path, and move, without waiting for perfect specs.
• Comfortable with non-determinism: LLMs don't return the same thing twice, and you're at home reasoning about failure modes, evals, and guardrails.
• Product-minded: your architecture and model choices trace back to real value for the people using the system.
• A clear communicator: you can explain an agent pipeline to a CTO or to a non-technical domain expert, and both come away confident.

What You'll Do

• Design and build the agent orchestration layer, including workflow state, event-driven agent execution, and next-best-action logic.
• Integrate LLMs behind a provider-agnostic layer (Anthropic, OpenAI, Writer.ai) with structured outputs, tool-calling, retries, and model routing.
• Turn messy inputs like documents, transcripts, and records into structured, reliable data the system can act on.
• Build human-in-the-loop approval flows that keep people in control of the decisions that matter.
• Set up evaluation and tuning loops so agent behavior improves on evidence rather than guesswork.
• Collaborate with frontend, product, and client stakeholders to ship end to end.

What You Bring

• Hands-on experience building LLM-powered or agentic applications, including tool-calling, structured output, RAG, and prompt iteration.
• 5+ years shipping production software. We build in TypeScript / Node, and we care more about your depth with AI systems than the language on your resume, so strong engineers from other ecosystems (Python and others) are welcome.
• Comfort with an orchestration framework (e.g. LangGraph.js, Vercel AI SDK, Mastra) and provider SDKs (Anthropic, OpenAI).
• Experience designing scalable systems on AWS.
• A track record of working with unstructured data: parsing, extraction, and validation.
• An eval-driven mindset, where you measure agent quality and iterate accordingly.
• Strong English communication skills.
• Comfortable in a fast-paced, agile consulting environment.

Extra Credit

• You've shipped AI into a regulated or high-stakes domain such as legal, health, or fintech.
• You've built eval harnesses for LLM systems rather than relying on manual spot checks.
• You've worked on voice or transcript pipelines, or document-heavy workflows.
• You know your way around Next.js and can meet the frontend halfway.
• Writer.ai experience, or open-source agent work you're proud to link.

#LI-PP1

#LI-REMOTE