GenAI · Agentic AI · RAG

Md Manawar Iqbal

AI/ML engineer with ~4 years of experience building production-oriented LLM applications, RAG pipelines and agentic systems. I turn messy business problems into reliable AI workflows, from classical ML and PyTorch models to RAG and multi-agent systems. Based in Pune, India.

See my work GitHub LinkedIn X

Projects

Enterprise

Agentic Incident Management

Architected an enterprise platform where specialized agents handle incident understanding, knowledge retrieval, assignment, resolution generation, validation and human review.

LangGraphHugging FaceRAGFastAPIDockerKubernetes
Open source

DataPilot

An autonomous data scientist for tabular data. The LLM decides what to investigate; deterministic Python tools produce the evidence. Runs fully local.

LangGraphGemma 3 4BOllamaTabPFNStreamlit
Hackathon · in progress

IntentGuard

An MCP proxy that checks every agent tool call against the user's original intent, built for ArmorIQ's Flow State hackathon.

MCPAgent safetyPython
Hacktoberfest

Touch Grass Walking Companion

A screen-free, voice-first companion that picks a loop route, narrates plants, birds and local history, and logs what you saw.

WhisperGemmaElevenLabs

Experience

Software Engineer – AI/ML · HCLTech, Noida

Sep 2022 – Present
  • Built production-grade ML, deep learning, GenAI and agentic AI solutions for enterprise applications.
  • Developed ML pipelines covering preprocessing, feature engineering, cross-validation, hyperparameter tuning and evaluation with Scikit-learn, plus ANN, CNN, RNN, LSTM, GRU and Transformer models in PyTorch.
  • Architected an agentic incident-management platform with LangGraph and Hugging Face, orchestrating specialized agents with a human review step.
  • Built RAG pipelines: ingestion, chunking, embeddings, vector indexing, hybrid retrieval, metadata filtering, reranking and query rewriting.
  • Implemented LLM evaluation with golden datasets, LLM-as-a-Judge, faithfulness, relevance and context precision/recall.
  • Integrated tool calling and MCP with enterprise APIs and databases; shipped FastAPI microservices on Docker, Kubernetes, CI/CD and AWS.
  • Added production observability: structured logging, metrics, tracing, latency monitoring, retries, timeouts, fallbacks and health checks.

B.Tech, Computer Science & Engineering · MAKAUT

2018 – 2022
  • Microsoft Azure Fundamentals (AZ-900)

Skills

GenAI & LLM

LLMsRAGEmbeddingsHybrid searchRerankingQuery rewritingPrompt/context engineeringFunction callingFine-tuning

Agentic AI

LangGraphLangChainAutoGenSemantic KernelMulti-agentReActMCPHITLHugging Face

LLM Evaluation

LLM-as-a-JudgeGolden datasetsRAG evaluationFaithfulnessRelevanceContext precision/recallAnswer correctness

Machine Learning

Scikit-learnNumPyPandasSciPyXGBoostRandom ForestSVMK-MeansPCAFeature engineering

Deep Learning

PyTorchANNCNNRNNLSTMGRUTransformersTransfer learningLoRA/PEFT

Backend & Cloud

PythonSQLFastAPIPydanticPostgreSQLRedisChromaDBFAISSAzure AI SearchAzure AI FoundryAWSDockerKubernetesGitHub ActionsCI/CD

Let's talk

Interested in Agentic AI, RAG, document intelligence and production AI systems. Reach out about roles or collaborations.

mmiqbalmanawar@gmail.com LinkedIn GitHub