my projects.

Passionate about building high-performance and scalable AI models. I thrive on solving complex problems, optimizing performance, and creating intuitive user experiences. My expertise lies in AI, Machine Learning, and Deep Learning, with a strong focus on delivering quality solutions.

Skillcheck

  • Evidence-backed static analysis scanner for AI-agent skill packages, categorizing findings as MALICIOUS, DANGEROUS, SUSPICIOUS, UNVERIFIED, or NO_FINDINGS, each backed by quoted evidence.
  • Parses Markdown structure, recursively decodes obfuscated payloads (base64/hex/rot13), and runs detectors across prose, Python AST, shell commands, secrets, and Unicode tricks — with multilingual prompt-injection detection across 8 languages.
  • FastAPI
  • SQLite
  • Docker
  • Security
  • Static Analysis

Slm From Scratch

Confidential / Internal
  • Built a modular small language model architecture from the ground up, implementing Grouped Query Attention (GQA), SwiGLU activation functions, Rotary Positional Embeddings (RoPE), and RMSNorm layers.
  • Engineered custom variants including a compact 4M-parameter token generator and a Qwen2.5-micro replication to study compute-efficient model convergence constraints.
  • PyTorch
  • Python
  • Transformers
  • LLM

Transformers From Scratch

Confidential / Internal
  • Developed the complete vanilla Transformer architecture ("Attention Is All You Need") from raw mathematical specifications, without relying on high-level transformer packages.
  • Built custom scaled dot-product attention units, token embedding steps, static positional encodings, and optimization loops to analyze encoder-decoder layer mechanics during training.
  • PyTorch
  • NumPy
  • Deep Learning

Paperly

  • Helps people understand research papers and converts them into notebooks for implementation.
  • Streamlines the process of moving from academic research to practical code implementation.
  • AI/ML
  • Research
  • Notebooks

Docuflow

  • Converts markdown files to docx documents seamlessy.
  • Simple and efficient tool for documentation conversion workflows.
  • Markdown
  • Converter
  • Utilities

Eco Sentry

  • Engineered an early wildfire prediction system for Indian forests by analyzing satellite signals and generating Convolutional Neural Network (CNN) heatmaps.
  • Achieved 98% accuracy in early detection (within 15 minutes) by processing real-world MODIS thermal anomaly datasets spanning 2015 to 2023.
  • Python
  • AI/ML
  • GEE
  • CNN
  • Satellite Data

Atithi Rakshak

  • Tourist safety app enabling contact even in no-network zones via mesh-style relays.
  • 99.9% message delivery in dead zones using LoRa protocol and offline relay architecture.
  • Full Stack
  • IoT
  • Python
  • C / C++
  • Offline Comms

Traffic Simulator

  • Route visualization and traffic simulation tuned for dense Indian road networks.
  • Implemented pathfinding algorithms and real-time data visualization for urban traffic prediction.
  • Python
  • Algorithms
  • Pathfinding
  • Data Viz

Block Coder

  • Competitive coding 2v2 multiplayer game with AI-powered Solidity code generation.
  • Reduced smart contract dev time by 40% with LLM tooling; Security Score 9.5/10.
  • React
  • Solidity
  • AI
  • LLM Agents
  • Multiplayer

Chain Sentinel

  • Flags malicious transactions from mempool behavior graphs to keep wallets safe.
  • Ensured $20k+ cost savings every 15 mins through proactive issue detection in beta testing.
  • FastAPI
  • Web3
  • Anomaly Detection
  • Security
  • Python

Cafe Management

  • My first self-coded project — lightweight management system for a small cafe.
  • Built on Python with inventory tracking and billing features.
  • Python
  • Inventory
  • Billing

Let's work together.

I'm always interested in new opportunities and exciting projects. Whether you have a project in mind or just want to chat about tech, I'd love to hear from you.

Currently available for freelance work and internship opportunities

Response time: Usually within 24 hours