Role Summary
We are seeking an Agentic AI Developer/Consultant who is fluent in the emerging world of agent-based AI systems — combining deep understanding of LLM orchestration, tool-augmented reasoning, autonomous agent frameworks, and real-world task automation.
This role is not for AI generalists or weekend hackers. We're looking for someone who lives and breathes structured autonomy, has built (not just read about) systems like AutoGPT, LangGraph, or ReAct-based agents, and understands both the promise and the pitfalls of autonomous execution layers.
Responsibilities
- Design, build, and deploy LLM-powered autonomous agents capable of reasoning, tool use, and long-horizon goal execution.
- Architect multi-agent systems that work in coordination (think swarms, workflows, or parallelized task execution).
- Integrate LLMs (GPT-4, Claude, Mistral, Gemini, etc.) with external tools, APIs, databases, and reasoning engines.
- Implement memory modules, retrieval-augmented generation (RAG), dynamic prompt orchestration, and self-evaluation protocols.
- Drive PoCs to production-grade systems with real-world constraints: latency, auditability, scale, and failure tolerance.
- Collaborate closely with our AI Strategy and Platform teams to define reusable patterns, guardrails, and agent behaviors.
- Stay ahead of the curve — R&D is not optional here. You're expected to anticipate what's next in the space and act on it.
Requirements
- 8+ years in software engineering or AI system development, with at least 2+ years in LLMs, agents, or similar paradigms.
- Deep experience with LangChain, LangGraph, CrewAI, AutoGen, Semantic Kernel, or custom-built agentic stacks.
- Proficiency in Python, with strong software design fundamentals (async, microservices, cloud-native workflows).
- Proven ability to build production-grade AI systems — not just demos.
- Understanding of task decomposition, feedback loops, contextual memory, and prompt engineering at scale.
- Bonus: Exposure to real-world deployments in fintech, insurtech, digital operations, or enterprise process automation.
What Success Looks Like
- You can take a user story and architect a thinking, reasoning system that executes it with minimal human intervention.
- You don't just follow the agentic AI playbook — you write new pages.
- You deliver value, not hype. Output that ships, scales, and solves problems.
What We Offer
- A high-autonomy, high-accountability culture.
- Early mover advantage in agentic systems within enterprise contexts.
- Direct access to leadership, sharp peers, and meaningful problems.
- Competitive compensation, high-growth trajectory, and global exposure.
