AI Agentic System Architect

About the Role:

As our Agentic System Architect, you will define and own the end-to-end architecture of our Python-based autonomous agent platform. Leveraging cutting-edge frameworks—LangChain, LangGraph, RAG pipelines, and more—you’ll ensure our multi-agent workflows are resilient, scalable, and aligned with business objectives

Key Responsibilities

Architectural Strategy & Standards

• Define system topology: microservices, agent clusters, RAG retrieval layers, and knowledge-graph integrations.
• Establish architectural patterns for chain-based vs. graph-based vs. retrieval-augmented workflows.

Component & Interface Design

• Specify Python modules for LLM integration, RAG connectors (Haystack, LlamaIndex), vector store adapters, and policy engines.
• Design REST/gRPC and message-queue interfaces compatible with Kafka/RabbitMQ, Semantic Kernel, and external APIs.

Scalability & Reliability

• Architect auto-scaling of Python agents on Kubernetes/EKS (including GPU-enabled inference pods).
• Define fault-tolerance patterns (circuit breakers, retries, bulkheads) and lead chaos-testing of agent clusters.

Security & Governance

• Embed authentication/authorization in agent flows (OIDC, OAuth2) and secure data retrieval (encrypted vector stores).
• Implement governance: prompt auditing, model-version control, drift detection, and usage quotas.

Performance & Cost Optimization

• Specify profiling/tracing requirements (OpenTelemetry in Python) across chain, graph, and RAG pipelines.
• Architect caching layers and GPU/CPU resource policies to minimize inference latency and cost.

Cross-Functional Leadership

• Collaborate with AI research, DevOps, and product teams to align roadmaps with strategic goals.
• Review and enforce best practices in Python code, CI/CD (GitHub Actions), and IaC (Terraform). 7. Documentation & Evangelism
• Produce architecture diagrams, decision records, and runbooks illustrating agentic designs (ReAct, CoT, RAG).
• Mentor engineers on agentic patterns—chain-of-thought, graph traversals, retrieval loops—and Python best practices.

Preferred Qualifications

Bachelor’s Degree in Computer Science, Information Technology, or related fields

(e.g., B.Tech, B.E., B.Sc. in Computer Science)

Preferred/Ideal Educational Qualification:

• Master’s Degree (optional but highly valued)

  in one of the following:
• M.Tech or M.E. in Computer Science / AI / Data Science
• M.Sc. in Artificial Intelligence or Machine Learning
• Integrated M.Tech programs in AI/ML from top-tier institutions like IITs, IIIT-H, IISc

• Ph.D. or Research Experience

  in NLP, Information Retrieval, or Agentic AI (especially relevant if applying to R&D-heavy teams like Microsoft Research, TCS Research, or AI startups)
• Certifications or online credentials in:
• LangChain, RAG architectures (DeepLearning.AI, Cohere, etc.)
• Advanced Python (Coursera/edX/Springboard/NPTEL)
• Cloud-based ML operations (AWS/Azure/GCP)

Additional Skill Set:

• Hands-on with agentic frameworks: LangChain, LangGraph, Microsoft AutoGen
• Experience building RAG pipelines with Haystack, LlamaIndex, or custom retrieval modules
• Familiarity with vector databases (FAISS, Pinecone, Chroma) and knowledge-graph stores (Neo4j)
• Expertise in observability stacks (Prometheus, Grafana, OpenTelemetry)
• Background in LLM SDKs (OpenAI, Anthropic) and function-calling paradigms Core Skills & Competencies
• System Thinking: Decompose complex business goals into modular, maintainable components
• Python Mastery: Idiomatic Python, async/await, package management (Poetry/venv)
• Distributed Design: Microservices, agent clusters, RAG retrieval loops, event streams
• Security-First: Embed authentication, authorization, and auditabilitys
• Leadership: Communicate complex system designs clearly to both technical and non-technical stakeholders

We are looking for someone with a proven track record in leveraging cuing-edge agentic frameworks and protocols. This includes hands-on experience with technologies such as Agent-to-Agent (A2A) communication protocols, LangGraph, LangChain, CrewAI, and other similar multi-agent orchestration tools. Your expertise will be crucial in transforming traditional, reactive AI applications into proactive, goal-driven intelligent agents that can signicantly enhance operational eciency, decision-making, and customer engagement in high-stakes domains. We envision this role as instrumental in driving innovation, translating cuing-edge academic research into deployable solutions, and contributing to the development of robust, scalable, and ethical AI agentic systems.