Expert Python Developer AI Platform & Agentic Systems

Location : Chennai / Pune / Remote

Experience : 12+ Years in Python

Budget : Open for the right candidates

About the Role :

We are looking for an Expert Python Developer to lead the engineering and technical implementation of an Agentic AI Platform's orchestration layer. The role involves building highly scalable backend services, lightweight Python SDKs, multi-tenant APIs, and secure execution environments for autonomous AI agents. The ideal candidate should possess deep expertise in Python, FastAPI, workflow orchestration, cloud-native architecture, and AI agent frameworks. The individual will work closely with DevSecOps, Data Engineering, and Product teams to design secure, resilient, and cost-efficient AI platforms.

Roles & Responsibilities :

Platform Control Plane & AI Gateway :

- Architect and maintain platform Control Plane and API Gateway using FastAPI.

- Build services for LLM routing, telemetry collection, cost metering, and request management.

- Design provider failover and routing strategies across multiple LLM providers.

Agentic SDK Engineering :

- Develop and maintain Python SDKs that abstract complex AI workflows.

- Build standardized interfaces for Human-in-the-Loop (HITL) workflows and agent overrides.

- Design reusable client libraries for engineering teams.

Workflow Orchestration :

- Implement long-running, fault-tolerant workflows using Temporal.

- Design multi-stage agent execution pipelines and recovery mechanisms.

- Build workflow observability and monitoring capabilities.

Multi-Tenant Backend Services :

- Design scalable RESTful APIs and backend microservices.

- Implement tenant isolation, data segregation, and secure CRUD operations.

- Develop APIs for administration, configuration management, and agent lifecycle management.

Infrastructure & Security :

- Collaborate with DevSecOps teams to deploy applications on Kubernetes.

- Integrate enterprise identity solutions using WorkOS.

- Implement Role-Based Access Control (RBAC), SSO, SCIM, and OAuth 2.0 security models.

Data Architecture & Agent Data Access :

- Integrate PostgreSQL, Neo4j, Graph databases, and Vector databases.

- Implement Model Context Protocol (MCP) to secure data access patterns.

- Design schema evolution and secure data contracts for AI agents.

LLM Gateway & Cost Governance :

- Build LiteLLM gateway capabilities including :

1. Provider routing

2. Rate limiting

3. Token budget management

4. Failover mechanisms

5. Cost tracking

6. Runaway loop protection

Production Operations :

- Define SLOs, observability standards, and operational metrics.

- Participate in incident response management and disaster recovery planning.

- Design systems for RTO/RPO requirements.

Data & Ingestion Pipelines :

- Build document ingestion pipelines using S3, queues, and OCR integrations.

- Design data pipelines integrating with existing data warehouses and Java/Spring Boot applications.

- Manage schema versioning and secure agent data consumption.

Identity & Access Management :

- Implement agent identity models and On-Behalf-Of (OBO) workflows.

- Build enterprise-grade audit logging and attribution mechanisms.

- Integrate SSO and SCIM provisioning for enterprise customers.

Required Experience :

- 10+ years of software engineering experience with strong Python expertise.

- Extensive experience building production-grade REST APIs and microservices using FastAPI.

- Hands-on experience with LangChain and LangSmith.

- Experience building multi-step reasoning systems and autonomous agents.

- Strong AWS experience including Bedrock, ECS, and S3.

- Hands-on experience with Docker and Kubernetes.

- Experience implementing workflow orchestration using Temporal.

- Experience with event-driven architectures using RabbitMQ or AWS SQS.

- Expertise in OAuth 2.0 and enterprise identity integrations.

- Experience with PostgreSQL, Neo4j/GraphDB, and Vector Databases.

- Experience designing versioned APIs and reusable SDKs.

Preferred Experience :

- Knowledge of Firecracker MicroVMs.

- Familiarity with Prodigy Framework.

- Experience building AI evaluation frameworks and hallucination measurement systems.

- Exposure to Commercial Real Estate (CRE) workflows.