Lead Architect – AI Systems & Cloud Infrastructure
Company: Humigent AI
Location: Gurgaon
Experience: 10+ years
Role Type: Full-Time | Leadership Track
Domains: Agentic AI, Multi-Cloud DevOps, Knowledge Graphs, Scalable Enterprise Applications
Role Overview
The Senior Architect – DevOps & Full-Stack will:
- Own the architecture, scalability, reliability, and security of Humigent AI’s applications and agentic systems
- Lead multi-cloud deployment across AWS + Azure
- Architect knowledge-graph-driven AI applications
- Mentor engineering teams and build a world-class technical foundation for all Humigent AI products
This role is a hands-on senior technical leadership position, combining architecture, DevOps excellence, full-stack engineering, and platform thinking.
Key Responsibilities
- Architecture Leadership
- Define architecture for Agentic AI applications using LangGraph + LLMs + Knowledge Graphs
- Design scalable microservices, event-driven systems, and multi-agent execution environments
- Establish architecture patterns for ingestion, retrieval, reasoning, evaluation, and UI layers
- Create end-to-end diagrams, design docs, standards, and reusable platform components
- Multi-Cloud DevOps & Infra (AWS + Azure)
Mandatory expertise in both clouds
- Architect and manage production workloads on:
- AWS: EC2, ECS/EKS, Lambda, S3, RDS, DynamoDB, API Gateway
- Azure: App Service, Functions, Azure OpenAI, Blob Storage, Azure SQL/PostgreSQL
- Build secure VPC/VNet architectures
- Implement IaC using Terraform / CloudFormation
- Establish logging, monitoring, and observability (CloudWatch, Azure Monitor, ELK, Prometheus, Grafana)
- Create CI/CD pipelines with GitHub Actions and Docker
- Lead performance engineering, autoscaling, load balancing, and failover strategies
- Full-Stack Engineering (Hands-On)
- Architect and review backend systems using Python (FastAPI/Flask)
- Drive front-end architecture using React/Next.js
- Create high-performance APIs, reusable components, and scalable microservices
- Own versioning, releases, code quality, and branching strategy
- Knowledge Graph + Ontology Architecture
Mandatory deep experience
- Architect solutions using Neo4j, AWS Neptune, Azure Cosmos DB (Gremlin)
- Define ontology schemas (OWL/JSON-LD), triple models, semantic layers
- Build reasoning pipelines (GraphRAG, hybrid KG + LLM retrieval)
- Integrate KGs with multi-agent systems for deterministic insights
- Ensure graph performance, security, versioning, and governance
- LLM, RAG & Agentic Integration
- Architect RAG pipelines with embeddings, vector DBs, caching, and re-ranking
- Integrate with OpenAI, Azure OpenAI, AWS Bedrock
- Build LangGraph, CrewAI and other Agentic AI based flows with custom tools and evaluators
- Define memory strategies, state machines, and agent orchestration patterns
- Leadership, Mentorship & Process
- Mentor full-stack, backend, DevOps, and agentic engineers
- Establish coding guidelines, architecture reviews, and best practices
- Collaborate with founders, product teams, domain experts, and clients
- Translate business use cases into scalable system designs
Required Skills (Mandatory)
Technical Mastery
- Python (FastAPI/Flask) + React/Next.js
- AWS + Azure cloud architecture
- Docker, Kubernetes, GitHub CI/CD
- Knowledge Graph technologies (Neo4j/Neptune/CosmosDB)
- Ontology modeling (OWL, RDF, JSON-LD)
- Vector DBs: Pinecone, Qdrant, Chroma, Weaviate
- LangGraph or CrewAI multi-agent system design
- SQL mastery (PostgreSQL preferred)
- Ingestion pipelines (structured + unstructured data)
Soft Skills
- Strong ownership and architectural thinking
- Excellent communication; ability to influence and guide teams
- Ability to simplify complex concepts for business stakeholders
- Calm under pressure; high accountability
- Passion for AI, cognitive reasoning, and building world-class systems
Nice to Have
- Experience in pharma, healthcare, food-tech domains
- Experience with metric layers (DBT, semantic layers, metric stores)
- Experience with time-series modeling and forecasting
- Publications, open-source contributions, patents