Agentic AI Engineer

Multiscale AI; India; On-site; INR 4,000,000 - 7,000,000; Full time

Benefits offered by this job

On-site in Hyderabad

Remote possible for some roles

Competitive compensation

Job summary

Multiscale AI is building the world’s first vertical-AI platform for microelectronics. You will lead the architecture and scaling of agentic systems that design experiments, optimize manufacturing steps, and accelerate scientific discovery, working across multi-agent orchestration, security, and deployment patterns.

Join a small team focused on reliable production-grade AI in a semiconductor context, with on-site opportunities in Hyderabad and potential remote considerations for some roles.

Qualifications

• 3–5 years building production AI/ML systems.
• Experience with LangGraph or equivalent orchestration.
• Knowledge of agent failover and production resilience.
• Experience with RBAC/security patterns.
• Experience with cloud deployment and containerization.

Responsibilities

• Architect multi-agent orchestration layers for production use.
• Develop MCP server and tool registration APIs for persistent agent context.
• Implement enterprise security primitives: RBAC, auth passthrough, audit trails.
• Provide network-isolated LLM serving for air-gapped environments.
• Design and optimize RAG pipelines and vector store architectures.

Skills

LangGraph experience

Distributed systems

MLOps

Cloud deployment

Tools

LangGraph

Kubernetes

Docker

GCP/AWS/Azure

Job description

Multiscale AI provides AI Agents and Digital Twins for Semiconductor Manufacturing. With offices in Atlanta, San Francisco, and Hyderabad, we help semiconductor manufacturers optimize fabs and achieve measurable results.

The Role

We're building the world's first vertical-AI platform for microelectronics and materials science. Our platform powers autonomous AI agents that design experiments, optimize manufacturing steps, and accelerate scientific discovery. You'll lead the charge in architecting and scaling these agentic systems.

What You'll Build

• Multi-agent orchestration layers using LangGraph - deterministic, observable, and built to survive real production conditions


• MCP server and tool registration APIs that give agents persistent context and shared state across complex workflows


• Enterprise security primitives: RBAC for agents, authentication passthrough, user impersonation, and immutable audit trails


• Network-isolated LLM serving for air-gapped semiconductor customer environments


• RAG pipelines and vector store architectures tuned for materials science and fab domain retrieval


• CI/CD pipelines and zero-downtime deployment patterns that keep production systems reliable


• The API contracts with the core Multiscale AI platform - the foundation everything else depends on


• Sandboxed, per-session compute for agents - isolated containers, orchestration and scheduling of long-running agent workloads


• Durable agent state and event streams - checkpointing, resumable runs, and auditable histories of every agent action


• Streaming chat APIs with reconnection, plus monitoring and evaluation of agents in production

What We're Looking For

• 3-5 years building production AI/ML systems - shipped and running, not demo-grade


• Hands-on with LangGraph or equivalent; you have opinions on orchestration frameworks and can defend them


• You've thought hard about how agents fail in production and built systems that recover gracefully


• Experience with enterprise security patterns: RBAC, auth passthrough, audit logging, or SOC 2 compliance


• Comfortable deploying on GCP (Vertex AI), AWS, or Azure - Docker and Kubernetes are second nature


• You understand RAG pipelines end to end - embeddings, retrieval, evaluation, and where they break

Nice to Have

• Any exposure to semiconductor manufacturing, fab operations, or materials science


• Experience with network-isolated or air-gapped deployment environments


• Open-source work or side projects in agentic AI - we want to see what you've built

Why Join Us

• You'll architect systems that don't exist yet - real ownership from day one


• Small team, fast decisions, direct access to customers and their hardest problems


• Work at the intersection of frontier AI and one of the world's most complex engineering domains


• Competitive compensation

Location: Hyderabad, India. Full-time. Preferred ON SITE. REMOTE can be considered for some positions