Hi, I'm Suman

|

AI Engineer specializing in Agentic Systems, RAG pipelines, and production-grade LLM applications.

I design and deploy scalable AI systems that automate complex workflows using multi-agent architectures, vector databases, and cloud-native infrastructure.

Built systems handling 19K+ images, deployed AI pipelines with Docker & CI/CD, and developing real-time agentic AI platforms.

Open to AI Engineer / GenAI roles
Flagship Project

Autonomous Multi-Agent AI System for Enterprise Knowledge Automation

Problem

Enterprises struggle to extract actionable insights from unstructured data. Manual research takes days, is prone to bias, and can't scale across domains.

Solution

Built a multi-agent system using RAG + LLM orchestration. Specialized AI agents collaborate autonomously to gather, analyze, cross-reference, and synthesize information into comprehensive reports with citations.

FastAPILangGraphCrewAIChroma / FAISSPostgreSQLDockerStreamlit

90%

Reduction in manual research effort

360°

Multi-perspective analysis

<2min

Full report generation

System Architecture

User QueryLangGraph OrchestratorData GathererWeb scraping + APIsAnalyzerCross-reference + synthReport WriterCitations + PDF/DOCXRAG Pipeline (Chroma/FAISS)Structured ReportPostgreSQLDocker

What I Build

AI Systems for Real-World Problems

How I Build Systems

Architect

Design multi-agent architectures, RAG pipelines, and system topologies before writing code. Define agent roles, communication patterns, and failure modes upfront.

Build

Implement with FastAPI, LangGraph, and vector databases. Write testable, modular code with proper error handling, retry logic, and observability from day one.

Deploy

Containerize with Docker, deploy on AWS/Azure, set up CI/CD pipelines, and monitor with Prometheus + Grafana. Production means it works at 2 AM without you.

Typical System Stack

OrchestrationLangGraph / CrewAI / Custom
LLM LayerGPT-4 / Claude / Gemini / Local (Ollama)
RetrievalFAISS / ChromaDB / Pinecone + Embeddings
BackendFastAPI / Django + PostgreSQL / Redis
InfraDocker / AWS / Azure / CI-CD / Monitoring

Every project follows the same discipline: define the architecture, build modular components, deploy with monitoring, and iterate based on real-world performance data.

Engineering Insights

Open to AI Engineer / GenAI roles

Building AI systems that ship to production

I build multi-agent systems, RAG pipelines, and production AI applications. If you need an engineer who can architect, build, and deploy — let's talk.