class HarshitSingh:
def __init__(self):
self.role = "AI/ML Engineer & Systems Architect"
self.education = "B.Tech CSE @ BVDUCOE Pune (CGPA: 8.04)"
self.location = "Pune, Maharashtra, India"
self.current_focus = ["Deep Learning", "LLMs", "Distributed Systems"]
self.achievements = {
"hackathons": ["Smart India Hackathon 2024 Finalist",
"PAN-IIT Imagine Hackathon 2025 Finalist"],
"competitive_programming": {
"codeforces": "1200+",
"leetcode": "1600+"
},
"open_source": "CERN HSF & GSoC-Affiliated Contributor"
}
def get_current_work(self):
return [
" Contributing to CERN HSF scientific computing projects",
" Building LLM-powered automation systems",
" Researching advanced computer vision & NLP architectures",
" Designing high-performance distributed systems"
]
def get_expertise(self):
return {
"ai_ml": ["PyTorch", "TensorFlow", "Transformers", "OpenCV", "Scikit-learn"],
"backend": ["Node.js", "Next.js", "Microservices", "REST APIs", "WebSockets"],
"databases": ["MongoDB", "PostgreSQL", "Redis", "Firebase"],
"devops": ["Docker", "Linux", "Git", "CI/CD"],
"languages": ["Python", "C++", "JavaScript", "TypeScript", "Java"]
}| Achievement | Impact |
|---|---|
| Smart India Hackathon 2024 | 🥈 Finalist among 15,000+ teams nationwide |
| PAN-IIT Imagine Hackathon 2025 | 🥈 Finalist - Led cross-functional engineering team |
| Competitive Programming | 💻 Codeforces 1200+ • LeetCode 1600+ • Regular Contests |
| Production Systems | ⚡ Built systems processing 250+ jobs/day with 99% uptime |
🤖 SaaS Automation Builder - Production AI Workflow Engine
Event-driven multi-tenant automation platform with LLM-powered workflow generation
- Processes 250+ automated jobs per day with p99 latency <100ms
- LLM-assisted workflow creation reducing manual setup time by 40%
- Multi-tenant RBAC with PostgreSQL isolation across 15+ organizations
- Fault-tolerant execution engine achieving 99% successful job completion
- Built with:
Next.js 14BunClerk AuthNeon PostgresOpenAI APIStripe
Key ML Features:
- Natural language to workflow conversion using GPT-4
- Intelligent error detection and auto-recovery suggestions
- Predictive scheduling based on historical execution patterns
🔧 MiniGit - ML-Enhanced Version Control System
Lightweight distributed VCS with AI-powered commit intelligence
- Manages 20K+ objects with O(1) SHA-1 hash lookups
- Zlib compression achieving ~35% storage reduction via content deduplication
- Commit DAG supporting efficient traversal over 6K+ commits
- LLM-powered commit summarization reducing manual effort by 40%
- Built with:
C++File SystemsDAGsNLPOpenAI API
ML Integration:
- Automatic commit message generation from code diffs
- Semantic code change analysis using transformer models
- Intelligent merge conflict prediction
Distributed Rate Limiter - High-Performance Traffic Control
Redis-backed rate limiting system with ML-based anomaly detection
- Enforces limits on 4K+ requests/minute with O(1) token bucket validation
- Deployed across 3 microservices with p99 overhead <7ms
- Blocks 80%+ burst traffic during DDoS-like spikes
- ML-powered anomaly detection for traffic pattern analysis
- Built with:
Node.jsRedisDistributed SystemsTensorFlow
Advanced Features:
- Real-time traffic pattern learning using time-series analysis
- Adaptive rate limiting based on historical request patterns
- Automated threat detection with minimal false positives
Languages
Backend & Systems
🗄️ Databases
Frontend & Tools
|
|
| Platform | Rating | Badge |
|---|---|---|
| LeetCode | 1600+ |  |
| Codeforces | 1200+ |  |
| GitHub | Active Contributor |  |
** B.Tech in Computer Science & Engineering** • Expected 2027
Bharati Vidyapeeth (DU) College of Engineering, Pune • CGPA: 8.04/10
Remote • 2025
- Contributing to CERN's High-Energy Physics software stack using C++ and Python
- 10+ reviewed pull requests merged across 50K+ LOC scientific computing codebases
- Collaborating with international maintainers on performance optimization and feature development
- Following CI/CD best practices with automated testing and code review workflows
- Focus areas: Scientific computing, distributed systems, and performance-critical algorithms
const currentProjects = {
learning: [
" Advanced PyTorch techniques for production ML",
" Large Language Model fine-tuning & optimization",
" React Native for cross-platform AI applications",
" Generative AI architectures (GANs, Diffusion Models)"
],
building: [
" ISL (Indian Sign Language) Translation System using Computer Vision",
" Real-time object detection pipeline with YOLO & TensorRT",
" Custom transformer architecture for domain-specific NLP",
" Scalable microservices architecture with ML inference endpoints"
],
contributing: [
" CERN HSF scientific computing projects",
" Open source ML libraries and frameworks",
" Technical documentation for AI/ML tools"
]
};| Project | Technology Stack | Impact Metrics |
|---|---|---|
| SaaS Automation Platform | Next.js, LLMs, PostgreSQL | 250+ jobs/day, 99% uptime |
| MiniGit VCS | C++, NLP, File Systems | 20K+ objects, 35% storage saved |
| Rate Limiter Service | Node.js, Redis | 4K+ req/min, <7ms p99 latency |
| ISL Translation | PyTorch, OpenCV | Real-time inference at 30 FPS |
fun_facts = {
"motto": "I'm a proud jack of all trades — on a mission to master them all",
"coding_style": "Clean, efficient, and well-documented",
"debug_method": "Rubber duck debugging + strategic print statements",
"favorite_tech": "Anything that makes systems faster and smarter",
"current_obsession": "Building AI that actually solves real problems",
"ask_me_about": ["OpenCV", "Matplotlib", "System Design", "Competitive Programming"]
}