JAYKUMAR CHAUDHARI

About

AI/ML undergraduate building deep learning and NLP systems, including CNN-based image classification, transformer-based financial sentiment analysis, and on-device computer vision. Currently an Artificial Intelligence Intern at XtraGrad Pvt. Ltd., seeking AI/ML internship or data science roles.

Experience

• Artificial Intelligence Intern
  XtraGrad Pvt. Ltd.
  Jun 2026 – Present

  Selected via AICTE Government Portal.

• AI Developer
  Freelance
  Jun 2026 – 2026

  Solo AI project delivering personalized insights and recommendations based on air-quality conditions.

• AI Project Contributor
  AICTE IDE Bootcamp, EDII Ahmedabad
  Apr 2026 – 2026

  Contributed to ideation, solution development, and prototype presentation for TaxMind AI as part of a team project during the bootcamp.

Education

• B.Tech
  SVKM’s Institute of Technology, Dhule, Maharashtra
  Information Technology · 2023 – 2027

Skills

• SQL
• C
• Computer Vision
• Data Analysis
• Deep Learning
• Python
• Java
• Feature Engineering
• Matplotlib
• Model Evaluation
• NumPy
• Pandas
• scikit-learn
• OpenCV
• Flask
• PyTorch
• TensorFlow
• Seaborn
• Natural Language Processing
• Sentiment Analysis
• Explainable AI
• Hyperparameter Tuning
• Image Classification

Projects

• DamageTriage — Building Damage Classification System
  Python, PyTorch, ResNet18 (Siamese), Grad-CAM

  Built a Siamese ResNet18 building-damage classification pipeline on satellite imagery, comparing paired pre-/post-disaster images to assess damage severity. Added Grad-CAM heatmaps, low-confidence review queue, block-level damage aggregation, and automated PDF disaster reports.

• FaceShield Edge — Offline Facial Recognition & Liveness Detection
  Python, React Native, BlazeFace, MobileFaceNet (INT8)

  Built an on-device face verification and anti-spoofing pipeline deployed via React Native, achieving about 97.3% accuracy at sub-300ms latency with zero cloud dependency. Finalist, NHAI Hackathon 7.0.

• Finance-News-Impact-Simulator
  Python, NLP, Transformers, Pandas, Scikit-learn

  Built an NLP pipeline ingesting 500+ daily news signals with transformer-based sentiment classification, predicting directional market impact at about 76% accuracy. Engineered an interactive simulation dashboard covering sentiment trends, volatility correlation, and portfolio-level impact scores with cross-asset filtering.

Courses & certifications

• Oracle Cloud Infrastructure 2025 Certified AI Foundations Associate ↗ · Oracle · 2025
• GenAI Powered Data Analytics Job Simulation ↗ · Tata

🏆 Achievements & awards

• Finalist — Innovent 2025 · 2025

  Nationally competitive shortlist at NMIMS Sobus Innovation Council.

• Shortlisted — Amazon HackOn 2025 · 2025

  National-level competitive selection, Amazon India.

• Participant — HackWithMumbai 3.0 · 2025

  Microsoft Office Mumbai.