An agent on the telecom network gets asked for a subscriber's Wi-Fi password. The right technician gets it. An unauthorized queue doesn't. Context and identity decide in real time — that's the platform we're building.
We're looking for engineers who can own the architecture, not just implement it. You'll go deep on one or two of the following workstreams:
- Orchestration — multi-agent engines, reasoning loops, query planning/rewriting, short- and long-term memory
- Grounding — knowledge graphs, identity graphs, ontology mapping to telecom taxonomies, text-to-SQL over undocumented schemas, RAG over legacy PDFs
- Interop — A2A-compliant agent design, containerized MCP servers, third-party A2A integrations
- Trust — contextual guardrails, RBAC policy, tracing, and evals wired into CI
- Exploration — POCs on current frameworks, evaluated against the A2UI/AGUI landscape
Most of the work isn't model tuning — it's making legacy provisioning systems answer questions, writing evals that catch regressions, and debugging non-deterministic systems. The reasoning layer is the fun 20%.
Tech Stack: Python, GCP (ADK, Vertex AI, GKE),
Docker, REST microservices, LangChain, Langfuse, OpenTelemetry, Permit.io, DeepEval, GEval
Requirements
- Production-level Python experience
- Hands-on experience with real LLM systems in production (RAG, tool-calling agents, or orchestration)
- Docker and Kubernetes
- Strong SQL — able to design schemas, not just query them
- Willingness to learn telecom domain knowledge (takes weeks, not optional)
- Comfort working with immature tooling and debugging framework-level issues
How to Apply
Email
[email protected] with your resume, plus one of the following:
- A repo we can review, or
- A short paragraph on an agent system you built that broke, and what you changed to fix it
Questions about the role before applying are welcome at the same address.
Pay: ₹800,000.00 - ₹1,000,000.00 per year
Benefits:
- Commuter assistance
- Flexible schedule
- Health insurance
- Paid sick time
- Paid time off
- Work from home
Work Location: In person
📌 AI Engineer (Agentic AI Platform) — (India)
🏢 Eldridge Morgan
📍 India