Artificial Intelligence Engineer

AI Developer

(Pro-Code / AI Engineering Focus) — Corporate Platforms, M365 & AI/Automation

Location: Offshore (UK shift)

Role Summary

We are looking for a pro-code AI Developer who builds production-grade AI systems in code. You will design and build AI agents, RAG pipelines, and AI-powered applications primarily in Python — orchestrating LLMs with frameworks such as LangChain, LangGraph, and AutoGen, integrating them into enterprise systems via APIs and MCP, and deploying them on the cloud (including Azure AI Foundry). The ideal candidate has a strong software-engineering foundation, a strong bias to build, and can move from a loosely defined idea to a working proof-of-concept in days, not weeks.

What Youll Do

• Design and build AI agents, RAG pipelines, and AI-powered applications in code (primarily Python), orchestrating LLMs with frameworks such as LangChain, LangGraph, and AutoGen.
• Implement advanced RAG techniques — hybrid search, re-ranking, contextual chunking, and long-context strategies — and agentic orchestration patterns (ReAct, plan-and-execute, multi-agent).

• Build and integrate APIs and enterprise connectors, including MCP-based integrations, to connect AI capabilities with enterprise systems and data sources.
• Deploy, configure, and operate AI workloads on the cloud, owning solutions end-to-end from prototype to production.
• Design AI solutions that scale — accounting for performance, throughput, latency, cost, and reliability as usage grows from POC to enterprise scale.
• Rapidly prototype: take a loosely defined idea or business problem and produce a working proof-of-concept quickly, then iterate based on feedback.
• Apply sound engineering judgment — error handling, evaluation, observability, security, and access considerations — across the solutions you build.

Required Qualifications

• Strong hands-on software development experience in Python, with a solid engineering foundation (you can design, write, test, and debug production code).
• Hands-on experience with LangChain, LangGraph, and/or AutoGen, and with agentic orchestration patterns for building multi-agent or agentic AI systems.
• Proven experience building RAG systems, including advanced techniques such as hybrid search, re-ranking, contextual chunking, and long-context strategies.
• Experience designing and consuming APIs and integrations; familiarity with MCP (Model Context Protocol) as an emerging integration pattern.

• Looking to get Placed? Try our Placement Guarantee Plan

  Get Hired
  Strong, hands-on command of cloud concepts on at least one major provider (Azure preferred; AWS or GCP considered) — able to independently provision, configure, develop against, deploy, and operate AI workloads and cloud resources. Hands-on experience with Azure AI Foundry (or an equivalent managed AI platform).
• Understanding of how to design and scale AI solutions for production — performance, throughput, latency, cost optimization, and reliability — including patterns such as caching, asynchronous processing, load handling, and horizontal scaling.
• Demonstrated ability to rapidly prototype — translating an idea into a functional proof-of-concept quickly using AI-assisted development and fast iteration.

Preferred Qualifications

• Experience with AI evaluation and observability tooling (e.g., ragas, LangSmith, Promptflow evals, Azure Monitor).
• Familiarity with AI safety, responsible AI principles, and enterprise guardrail patterns (content filtering, grounding checks).
• Working knowledge of low-code AI platforms (Microsoft Copilot Studio, Power Platform) or agent-builder platforms (Lyzr, Moveworks) for rapid delivery where they are the right fit.
• Experience integrating with Microsoft 365 and Dataverse, or an equivalent enterprise ecosystem.

What Makes You a Fit

You are first and foremost a strong engineer who has moved into AI — comfortable living in code, reasoning about architecture, and owning a solution from prototype to production. You reach for low-code tools when they are genuinely the fastest path, but your default is to build in code, and you would rather ship a rough proof-of-concept today than a perfect specification next month.