Amarjeet Kumar profile photo 1 on Tal

Amarjeet Kumar

Aspiring Data Scientist | Machine Learning • Deep Learning • NLP • Computer Vision | Python, PyTorch, FastAPI | NIT Tiruchirappalli

India

< 1 year of work experience

National Institute of Technology, Tiruchirappalli
May 20253 mosJul 2025

Built an AI-driven autism screening pipeline using OpenCV and MediaPipe for gaze/facial-expression feature extraction; trained a custom CNN achieving 96% classification accuracy Optimized real-time landmark extraction (iris-ratio, facial pose estimation) for low-latency, non-invasive neurodevelopmental monitoring Skills tags: Python, OpenCV, MediaPipe, CNN, Deep Learning

Amarjeet’s AI Stack

antigravity
antigravity
geminicli
geminicli
github copilot
github copilot
hugging face
hugging face
perplexity
perplexity
Microsoft Copilot
Microsoft Copilot

Education

National Institute of Technology Tiruchirappalli (NIT Trichy)

Master of Science - MS, Computer Science

National Institute of Technology Tiruchirappalli (NIT Trichy)

20242 yrs2026

Bachelor of Science - BSc, Mathematics

Binod Bihari Mahto Koyelanchal University

20203 yrs2023

Join Amarjeet and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Amarjeet Kumar on Tal

Headline

Aspiring Data Scientist | Machine Learning • Deep Learning • NLP • Computer Vision | Python, PyTorch, FastAPI | NIT Tiruchirappalli

About

M.Sc. Computer Science student at NIT Tiruchirappalli, focused on applied AI — computer vision, NLP, and RAG systems that go beyond notebook demos into deployed, real-time products. Recent work: 🔹 Built an AI-driven autism screening pipeline (OpenCV + MediaPipe + custom CNN) hitting 96% classification accuracy, as a Research Intern at NIT Trichy 🔹 Designed a real-time multi-modal classroom attention system fusing MediaPipe FaceMesh, a ViT gaze regressor, and an emotion classifier into a live Streamlit dashboard with CSV export 🔹 Shipped a RAG pipeline with Qdrant vector search, OCR/web-scraping ingestion, and citation grounding — benchmarked chunking strategies to reduce hallucination 🔹 Architected and deployed a multi-provider AI gateway (FastAPI + React) routing between GPT-4o and Claude with SSE streaming, token-bucket rate limiting, and live usage analytics Tools: Python, PyTorch, TensorFlow, Hugging Face Transformers, LangChain, OpenCV, MediaPipe, FastAPI, Docker, Qdrant, MongoDB. 750+ DSA problems solved on LeetCode/GFG. NPTEL Elite certification in Privacy & Security in Online Social Media. Looking for full-time roles or internships in Data Science, ML Engineering, or Applied AI — open to relocation. Always happy to talk RAG, computer vision, or production ML.

Location

India

AI stack

antigravity

geminicli

github copilot

hugging face

perplexity

Microsoft Copilot

Work experience

Quantitative Researcher at National Institute of Technology, Tiruchirappalli, May 2025 - Jul 2025 · 3 mos, Built an AI-driven autism screening pipeline using OpenCV and MediaPipe for gaze/facial-expression feature extraction; trained a custom CNN achieving 96% classification accuracy Optimized real-time landmark extraction (iris-ratio, facial pose estimation) for low-latency, non-invasive neurodevelopmental monitoring Skills tags: Python, OpenCV, MediaPipe, CNN, Deep Learning

Education

Master of Science - MS, Computer Science, National Institute of Technology Tiruchirappalli (NIT Trichy)

Bachelor of Science - BSc, Mathematics, Binod Bihari Mahto Koyelanchal University

Amarjeet Kumar - Quantitative Researcher in India | Tal