Axia

Job Description: AI Solutions Architect / Team Lead Role: AI Solutions Architect / Team Lead Location: Pune (In Person) Type: Full-time Department: Engineering / AI Experience: 4-5 Years About the Role We are looking for a hands-on Solution Architect & Team Lead who knows that a great AI model is useless without a great API wrapping it. You won’t just be optimizing prompts; you will be architecting the distributed systems that serve them. You will bridge the gap between experimental AI and rigorous software engineering. You will lead a team to build production-grade, fault-tolerant systems where Clean Code, API standards, and System Design are just as important as the LLM context window. Key Responsibilities • System & API Architecture: Design high-performance, contract-first APIs (REST/gRPC/GraphQL). Define the interface between our AI agents and the frontend/client applications, ensuring low latency and type safety. • Production-Grade Engineering: Enforce software engineering best practices: CI/CD, unit/integration testing, code reviews, and design patterns (SOLID, DRY). Ensure the codebase is maintainable and modular, not a "script" thrown into production. • Agentic Systems & MCP: Design autonomous AI agents that can utilize tools via robust API integrations. Implement the Model Context Protocol (MCP) to standardize how models interface with external data and tools. • Retrieval (RAG) at Scale: Architect advanced RAG pipelines involving hybrid search and re-ranking, treating the vector database as a critical production component, not a sandbox experiment. • Team Leadership: Mentor engineers on both AI concepts and core backend development. Drive technical decisions on build vs. buy, schema design, and microservices architecture. • Security & Reliability: Implement rate limiting, circuit breakers, and rigorous authentication/authorization (OAuth/JWT) to protect expensive AI resources. Technical Requirements 1. Core Software & API Engineering (Essential) • Mastery of Backend Development: Expert-level proficiency in Python (FastAPI/Django) or Node.js/TypeScript. You write clean, typed, and self-documenting code. • API Design Excellence: Deep understanding of RESTful principles, gRPC (for low-latency internal comms), and GraphQL. Experience with OpenAPI/Swagger specifications is non-negotiable. • Database Design: Strong grasp of SQL (PostgreSQL) schema design and NoSQL patterns. You understand how to optimize queries for high-throughput applications. • Distributed Systems: Experience building event-driven architectures (Kafka/RabbitMQ/Redis) to handle asynchronous AI tasks and long-running agent processes. 2. Generative AI & Agents • Agentic Workflows: Proven track record building agents using frameworks like LangChain, LangGraph, or AutoGen. You understand state management in multi-turn conversations. • Model Context Protocol (MCP): Knowledge of MCP or similar protocols to manage how LLMs read/write context and interact with client-side applications. • LLM Integration: Experience orchestrating complex flows with OpenAI (GPT-4), Anthropic (Claude), and open-source models (Llama 3, Mistral). 3. MLOps & Production AI • Retrieval (RAG): Advanced experience with Vector Databases (Pinecone, Weaviate, Milvus). • Observability: Implementation of tracing and monitoring for AI chains (LangSmith, Arize, Datadog) to track latency, token usage, and errors. • Evaluation: Setting up "LLM-as-a-judge" pipelines to automatically verify API responses and model quality before deployment. 4. Cloud & Infrastructure • Cloud Native: Architecting on AWS/GCP/Azure. • Containerization: Docker and Kubernetes expertise. You know how to scale stateless API containers independently of stateful vector stores. • Infrastructure as Code: Terraform or CloudFormation. Success Metrics • System Stability: 99.9% API uptime and successful handling of rate limits/backpressure. • Code Quality: Reduction in technical debt and increased test coverage across the backend. • Performance: Optimized latency for AI responses through efficient API design and caching strategies. Why Join Us? • Engineering First: We don't just "wrap wrappers. " component, not a magic box. We build robust software where AI is a • Architectural Freedom: You define the stack, the API standards, and the protocols (like MCP) we use to scale. • High Impact: Your code will power the core intelligence of the product.