Skip to content
View Niket07pathak's full-sized avatar
👋
Hello
👋
Hello

Block or report Niket07pathak

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Niket07pathak/README.md

Typing SVG


LinkedIn GitHub Email AWS Dev Associate

⚡ What I'm About

I don't just train models — I build the systems around them.

MS CS grad from UIC (GPA 4.0) obsessed with the full AI stack: from raw data pipelines and model fine-tuning to deployed APIs and dashboards that actually work in production. I care about building things that are measurable, interpretable, and real.

Currently hunting for full-time AI/ML Engineer roles in the US.

🛠 Tech I Work With

Languages

Python SQL JavaScript C++ R

AI / LLM / ML

PyTorch HuggingFace LangChain OpenAI TensorFlow Scikit-learn MLflow

Backend / Frontend

FastAPI React Flask

Cloud / DevOps / Data

AWS Azure GCP Docker Kubernetes Kafka Spark

Vector Databases & LLM Infra

Pinecone ChromaDB LlamaIndex LangSmith

🔬 Featured Projects

🧠 Chain-of-Thought on CLEVR

Fine-tuned BLIP-2 OPT-2.7B with LoRA on 50K CLEVR samples using chain-of-thought supervision. Converted symbolic functional programs into natural-language CoT traces.

Result: +37pp accuracy over zero-shot baseline; +17pp CoT gain on short reasoning chains.

VLM LoRA PEFT Multimodal CoT

→ View Project

🦅 Wildlife Sound Monitoring System

Real-time edge AI system for live bird sound detection with sub-second streaming inference on NVIDIA Jetson GPU — with React dashboard, telemetry, and rare event alerts.

Result: ~40% reduction in dashboard latency via jtop-guided optimization.

Edge AI BirdNET FastAPI React WebSockets

→ View Project

⚖️ Fairness-Aware Credit Scoring

Benchmarked 10+ fairness-aware ML models across 4 real-world credit datasets. Introduced a Responsibility Score (RS) combining fairness + accuracy. Hybrid DI Remover + EG Reduction pipeline on HMDA data.

Result: 98% accuracy with LFR; fairness DI = 0.94; SHAP-validated interpretability.

Responsible AI Fairness SHAP ML HMDA

→ View Project

🤖 AI Project Manager (MCP)

Agentic AI project management system using MCP-based workflows, embeddings, and full-stack AI automation — bridging LLM reasoning with real project state.

Focus: MCP, agentic planning, task automation, full-stack AI.

MCP Agentic AI LangGraph Embeddings Full-Stack

→ View Project

🧭 Currently Focused On

◆ LLM Fine-tuning & Evaluation          ◆ RAG Pipelines & Vector Search
◆ Agentic AI / MCP Workflows            ◆ Edge AI & Real-time Inference
◆ MLOps & Cloud ML Pipelines            ◆ Responsible AI & Fairness Metrics
◆ FastAPI + React Full-stack Apps       ◆ Open to AI/ML Engineering Roles 🚀

🎓 Education & Certs
🏛️ MS Computer Science — University of Illinois Chicago GPA 4.0 · 2024–2026
🏛️ BTech Computer Science — SGSITS, Indore GPA 3.7 · 2020–2024
☁️ AWS Certified Developer — Associate Valid 2025–2028
☁️ AWS Certified Cloud Practitioner Valid 2025–2028

Let's build something worth deploying.

Profile Views

Pinned Loading

  1. Chain-of-Thought-on-CLEVR Chain-of-Thought-on-CLEVR Public

    Evaluating COT under supervision for VQA

    Python

  2. RAG_Adaptive_Pipeline RAG_Adaptive_Pipeline Public

    Jupyter Notebook

  3. Fairness-Aware-Credit-Scoring Fairness-Aware-Credit-Scoring Public

    A Comparative Study and a Hybrid Model Approach

    Jupyter Notebook

  4. LLM_BookReco LLM_BookReco Public

    Jupyter Notebook

  5. Wildfire-Analysis-of-the-United-States Wildfire-Analysis-of-the-United-States Public

    Jupyter Notebook

  6. student-notes-mcp student-notes-mcp Public

    Python