Agentic AI Systems
Multi-agent workflows, grounded retrieval, orchestration, tool use, validation, and business-safe response design.
Azure AI FoundryAzure OpenAILangGraphRAGPrompt Engineering
System live
Hi, I’mSatnam
Senior AI Engineer
I build grounded enterprise AI systems across Azure AI Foundry, Microsoft Fabric, MCP, FastAPI, document intelligence, and machine-learning workflows.
2.6
Years Experience
8+
Portfolio Projects
4+
Enterprise AI
System live
Hi, I’m Satnam
Senior AI Engineer
I build grounded enterprise AI systems across Azure AI Foundry, Microsoft Fabric, MCP, FastAPI, document intelligence, and machine-learning workflows.
“I connect reliable data, grounded reasoning, backend engineering, and validation into systems people can actually trust and use.”
2.6
Years Experience
8+
Portfolio Projects
4+
Enterprise AI Workstreams
Who I am
Senior AI Engineer
“I connect reliable data, grounded reasoning, backend engineering, and validation into systems people can actually trust and use.”
I’m Satnam, a Senior AI Engineer at ZS Associates. I work across the full enterprise AI stack — from governed data foundations and retrieval to orchestration, backend services, validation, and usable business output.
Read more Read less
I’m Satnam, a Senior AI Engineer at ZS Associates. I work across the full enterprise AI stack — from governed data foundations and retrieval to orchestration, backend services, validation, and usable business output.
The strongest systems are not the ones with the most prompts. They are the ones with clear contracts, reliable context, transparent validation, secure integration, and an experience that helps someone make a better decision.
Gurugram, India Senior AI Engineer Open to roles
Expertise
What I Build
The portfolio is organised around systems and outcomes, not a wall of disconnected technology badges.
Agentic AI Systems
Agentic AI Systems
Multi-agent workflows, grounded retrieval, orchestration, tool use, validation, and business-safe response design.
Azure AI FoundryAzure OpenAILangGraphRAGPrompt Engineering
MCP Knowledge Layers
MCP Knowledge Layers
Governed tool surfaces that expose product, KPI, schema, and semantic knowledge through a reusable contract.
MCPAzure AI SearchFunction AppsEntra IDManaged Identity
MCP Knowledge Layers
Governed tool surfaces that expose product, KPI, schema, and semantic knowledge through a reusable contract.
MCPAzure AI SearchFunction AppsEntra IDManaged Identity
Enterprise Analytics
Enterprise Analytics
Natural-language analytics over Microsoft Fabric with KPI planning, SQL generation, validation, charts, and insights.
Microsoft FabricSQL EndpointPySparkRedisDQM
Enterprise Analytics
Natural-language analytics over Microsoft Fabric with KPI planning, SQL generation, validation, charts, and insights.
Microsoft FabricSQL EndpointPySparkRedisDQM
Document Intelligence
Document Intelligence
Evidence-backed extraction, validation, and coauthoring workflows for complex documents and policy review.
PyMuPDFpdfplumberHybrid RetrievalGroqPydantic
Document Intelligence
Evidence-backed extraction, validation, and coauthoring workflows for complex documents and policy review.
PyMuPDFpdfplumberHybrid RetrievalGroqPydantic
Backend & Cloud
Backend & Cloud
Typed FastAPI services, asynchronous orchestration, serverless Azure integration, containerization, and observability.
PythonFastAPIAsyncIOACADockerAzure Monitor
Backend & Cloud
Typed FastAPI services, asynchronous orchestration, serverless Azure integration, containerization, and observability.
PythonFastAPIAsyncIOACADockerAzure Monitor
Machine Learning
Machine Learning
Classical ML, recommender systems, similarity search, NLP inference, feature engineering, and model-backed applications.
scikit-learnCosine SimilarityNLPPandasNumPy
Machine Learning
Classical ML, recommender systems, similarity search, NLP inference, feature engineering, and model-backed applications.
scikit-learnCosine SimilarityNLPPandasNumPy
Experience
Work that built the foundation
Detailed enough to show ownership and progression, sanitized enough to protect confidential client work.
Building enterprise agentic AI platforms, MCP-enabled knowledge layers, document intelligence systems, FastAPI backends, Azure-native integrations, and Microsoft Fabric analytics workflows.
Key achievements
- Promoted after high-impact Agentic AI and GenAI project delivery across enterprise client workstreams
- Designed reusable multi-agent orchestration, bounded retrieval patterns, and decision-ready outputs
- Built backend contracts, validation loops, and enterprise integration flows
- + 1 more on desktop
Technologies
Azure AI FoundryMCPFastAPIMicrosoft FabricAzure AI SearchAKSRedis
Full time
2026 / NowSenior AI Engineer
ZS Associates Gurugram, India
Building enterprise agentic AI platforms, MCP-enabled knowledge layers, document intelligence systems, FastAPI backends, Azure-native integrations, and Microsoft Fabric analytics workflows.
Technologies
Azure AI FoundryMCPFastAPIMicrosoft FabricAzure AI SearchAKSRedis
Key achievements
- Promoted after high-impact Agentic AI and GenAI project delivery across enterprise client workstreams
- Designed reusable multi-agent orchestration, bounded retrieval patterns, and decision-ready outputs
- Built backend contracts, validation loops, and enterprise integration flows
- Recognised with the Client Impact Champion award for exceptional client delivery
Full time
2024 / 2025AI Engineer
ZS Associates Gurugram, India
Built data platforms, ETL and ELT pipelines, reusable data-quality accelerators, and backend services across Microsoft Fabric and Azure while leading technical delivery on sanitized client workstreams.
Technologies
PythonPySparkSQLDelta LakeLakehouseAzure MonitorFastAPI
Key achievements
- Worked on pipelines and analytics systems handling more than 120M records daily
- Created config-driven validation and monitoring patterns across Medallion lakehouse layers
- Built the data-engineering foundation and backend services for later agentic systems
Credentials
Education, certification & recognition
Flagship builds & portfolio
Proof that maps to the role
Flagship case studies and this interactive portfolio — built as a living technical demonstration. Each project page includes architecture, implementation details, and an interactive workflow trace.
Multi-Agent Orchestration · Enterprise Analytics
Enterprise Agentic Analytics Platform
A FastAPI-based multi-agent analytics platform that converts natural-language requests into governed KPI retrieval, query planning, Microsoft Fabric SQL execution, validation, charts, and structured business insights.
FastAPIAzure OpenAIMicrosoft Fabric+3
Multi-Agent Orchestration · Enterprise Analytics2025
Enterprise Agentic Analytics Platform
A FastAPI-based multi-agent analytics platform that converts natural-language requests into governed KPI retrieval, query planning, Microsoft Fabric SQL execution, validation, charts, and structured business insights.
Keywords
Multi-agent orchestration, Python backend ownership, analytical reliability, Microsoft Fabric integration, and typed response design.
- FastAPI analytics API with typed request and response contracts
- Redis-backed history, metadata, and warm connections
- Focused agents for intent, KPI planning, SQL, validation, and insight
- Azure AI Search retrieval over KPI and table metadata
FastAPIAzure OpenAIMicrosoft FabricAzure AI Search+2
Document Intelligence · Coauthoring
PolicyLens Document Intelligence Platform
PolicyLens extracts source-linked evidence from complex policy documents, validates structured outputs, and compresses a multi-day review and coauthoring process into a focused analytical workflow.
PythonPyMuPDFpdfplumber+3
Document Intelligence · Coauthoring2026
PolicyLens Document Intelligence Platform
PolicyLens extracts source-linked evidence from complex policy documents, validates structured outputs, and compresses a multi-day review and coauthoring process into a focused analytical workflow.
Keywords
Document intelligence architecture, constrained LLM extraction, evidence traceability, deterministic validation, business-rule execution, and workflow acceleration.
- Structured extraction for policy review and coauthoring
- Layered PDF parsing for text, layout, and tables
- Strict evidence bundles that keep every field traceable
- Hybrid retrieval for quickly isolating the right sections
PythonPyMuPDFpdfplumberHybrid Retrieval+2
Agentic AI · MCP
Enterprise Knowledge Layer & MCP Agent
An in-development enterprise knowledge layer that exposes curated business context through MCP, Azure AI Search, Azure AI Foundry agents, Azure Function Apps, and Microsoft Entra ID.
MCPAzure AI SearchAzure AI Foundry+2
Agentic AI · MCP2026
Enterprise Knowledge Layer & MCP Agent
An in-development enterprise knowledge layer that exposes curated business context through MCP, Azure AI Search, Azure AI Foundry agents, Azure Function Apps, and Microsoft Entra ID.
Keywords
MCP server design, enterprise retrieval, multi-index search, identity-aware Azure integration, and reusable knowledge-platform architecture.
- Remote MCP server hosted on Azure Function Apps
- Product registry resolving product-specific search indexes
- Azure AI Search indexes for KPI, schema, definitions, and relationships
- Focused Foundry agents for product, schema, and KPI identification
MCPAzure AI SearchAzure AI FoundryFunction Apps+1
Full-Stack Engineering · Agentic AI
Satnam's Interactive Agentic Portfolio
A full-stack interactive portfolio built with TanStack Start, Vite, React, Tailwind CSS, Groq Llama, Upstash Redis, and Vercel serverless APIs.
ReactTanStack StartVite+4
Full-Stack Engineering · Agentic AI2026
Satnam's Interactive Agentic Portfolio
A full-stack interactive portfolio built with TanStack Start, Vite, React, Tailwind CSS, Groq Llama, Upstash Redis, and Vercel serverless APIs.
Keywords
Frontend engineering, serverless API design, LLM orchestration, structured response handling, Redis-backed analytics, and end-to-end product delivery.
- CV-grounded Niki AI chat assistant with structured JSON responses and next-question follow-ups
- Inline schedule/booking system that emails visitor context and chat history directly to Satnam
- Realtime analytics modal tracking unique visitors, likes, countries, and devices via Upstash Redis
- Custom Vite dev middleware serving Vercel serverless API routes seamlessly during local development
ReactTanStack StartViteTailwind CSS+3
Build lab
Next systems on the roadmap
These are intentionally shown as planned builds, not completed projects.
Planned build 01
Veritas AI
Multimodal Content Verification System
A planned multimodal verification platform for text, images, video, audio, and screenshots. The system will return a trust score, risk level, and evidence-backed explanation rather than a binary label.
- Deepfake and manipulated-media signals
- Scam and phishing pattern detection
- RAG-backed claim verification
- Unified trust and risk scoring
PythonFastAPIOpenCVOCRNLPRAG
Planned build 02
GroundLoop RAG
Retrieval System with Closed-Loop Validation
A planned RAG pipeline that grades every generated answer against retrieved context, rewrites low-confidence queries, and retries instead of silently returning a confident hallucination.
- Per-response confidence scoring
- Automatic query rewriting
- Closed-loop retry
- Hybrid vector retrieval
LangGraphFAISSPineconeChromaDBLLM Grader
Niki AI
Ask or talk to the portfolio
Niki AI
Ask the portfolio directly
One grounded assistant, now available in text and live voice.
Satnam's AI Assistant
Niki AI
Grounded in Satnam's public CV, portfolio, and project context. Ask by text, talk to Niki live, or send Satnam a connection request.
Next step
Let’s talk about the role
If the work involves enterprise AI systems, agentic workflows, MCP, Microsoft Fabric, document intelligence, ML products, or backend architecture, a direct conversation is the best next step.
Emailsatnamsjob@gmail.comLinkedInlinkedin.com/in/22satnamGitHubgithub.com/22satnam
Hiring fit
- Agentic AI, GenAI, Python, and backend roles
- Azure AI Foundry, MCP, Fabric, search, and document intelligence
- India-based and open to the right engineering opportunity
- Comfortable discussing sanitized architecture and live demos