Senior AI Engineer, Pod Tech Lead (Bangalore)

👋 About Xanadu

Xanadu helps mid-market organisations in the Middle East move from AI experiments to AI that actually runs in production. We’re a Certified Services Partner, Claude Partner Network, building on Claude for the reasoning, workflow and agent layer, and a Canva partner for the creative and communications layer.

We work in pods: a Forward Deployed Engineer onsite with the customer in Dubai, and a small team in our build hub turning designs into agents that run. You lead the build side of a pod.

🧠 The role

You’re the technical lead of a pod. Part solution architect, part senior developer. You turn the FDE’s design into a build plan, write the hard parts yourself, review everything that ships, and own production quality.

🎯 What you’ll own

• Architecture. Work with the FDE and the pod’s AI Data Engineer on the reference architecture, data flows and integration design for each customer.
• The hard parts. Auth, orchestration, MCP servers, sub-agents. The bits that break demos.
• The hand-off. Own the line between the data side and the agent side: how the AI Data Engineer’s look-up tools and the AI Engineer’s agents fit together.
• Access. Set one approach for the pod to who sees what, so an agent only ever sees what its user is allowed to see. The AI Data Engineer and AI Engineer each build their side of it.
• Quality. Code review and test coverage across both sides of the pod. You confirm a build is technically ready; the FDE gives the final sign-off before it goes live.
• Production readiness. Security, monitoring, logging, runbooks, handover.
• Your people. Mentor the AI Engineer and AI Data Engineer in your pod, and the Associate shared across pods. Make them better, fast.
• Patterns. Pull reusable pieces out of every build and into our Agent Library.
• Customers. Join technical calls as the engineering counterpart when the FDE needs you.

📏 What good looks like

• 90 days: Claude-certified, leading the build on a live pod, first pattern contributed to the Agent Library.
• 12 months: tech lead on two or more production deployments, and a mid-level engineer you’ve grown toward senior.

🔑 Must-haves

• 5+ years in software engineering, including 1 to 2+ years shipping LLM applications to production (agents, tool use, retrieval, evals).
• Strong Python and/or TypeScript.
• APIs, cloud (AWS, Azure or GCP), CI/CD, containers.
• Enterprise integration scars: auth, SSO, messy data, brittle systems.
• Clear written English, and you can explain a trade-off to someone who doesn’t code.

🌟 Nice-to-haves

• Claude Agent SDK, Claude Code, MCP.
• Microsoft 365 and Graph, Salesforce, HubSpot or ERP integrations.
• Security thinking: threat modelling, least-privilege, data handling.
• You’ve led a small team before.
• Arabic-language or right-to-left data experience.

🚫 What this isn’t

• Not a research role. We don’t train models; we make them useful.
• Not a ticket queue. You’ll understand why you’re building something, and push back when it’s wrong.

🧪 How we hire

1. Intro call (30 min).
2. Values and behaviours (45 to 60 min).
3. Bar-raiser (60 min).
4. Paid working session (2 to 3 hours): pair-build a small Claude agent with tool use and evals, then talk us through how you’d take it to production.
5. Decision: within 48 hours of every stage.

💛 Why Xanadu

• Real production work for real customers across the Middle East.
• Direct line to the FDEs and the customer. You’ll know why the work matters.
• Certified on Claude on our dime, and a clear path to Build Hub Lead.
• Equity (ESOP), so you share in what you build.
• No egos, no passengers.