About the role

Staff AI Platform Engineer building Paytm Labs' inference and agentic AI platforms. Operating models and autonomous workflows across payments, risk, fraud, and enterprise fintech systems.

Responsibilities

Build and operate multi-model serving across text, voice, code, and vision modalities
Own the model lifecycle: download, deploy, serve, monitor, update, and swap models
Optimize inference latency, throughput, and cost through quantization, batching, caching, and routing strategies
Ensure inference reliability for agents and dependent systems
Architect and build the Agentic AI Platform, including runtime infrastructure, orchestration systems, and developer tooling
Design multi-agent coordination systems for complex workflows
Build secure tool-use infrastructure for APIs, databases, and services
Implement guarded workflow automation for multi-step business and engineering tasks
Build safety and guardrail systems, including permissioning, sandboxing, and human-in-the-loop workflows
Develop evaluation and observability frameworks for agent behavior, regression detection, and failure debugging
Develop SDKs and APIs for internal teams to build and deploy agents
Define technical direction and architecture for agentic systems across the organization
Establish patterns and standards for agent design, tool calling, and evaluation
Partner with ML, product, and security teams
Mentor engineers and contribute to agent system design best practices

8+ years of software engineering experience, with 3+ years in AI systems or LLM applications
Strong understanding of LLM-based agent architectures: tool use, multi-step workflows, multi-agent coordination, and failure modes
Experience building highly reliable distributed systems
Experience evaluating LLM systems in production, including building evals, detecting regressions, and debugging non-deterministic failures
Proficiency in TypeScript or Python, and willingness to work in both
Experience with modern LLM APIs or open-source models
Experience with or strong interest in model serving, including vLLM, TensorRT-LLM, or Triton
Understanding of distributed systems, including task queues, event-driven architectures, state management, and durable long-running workflows
Experience with AWS or GCP and containerized deployments
Strong understanding of security risks in agentic systems, including prompt injection, privilege escalation, and data leakage
Demonstrated experience leading complex technical initiatives
Strong written and verbal communication skills