koopatroopa787

Hey, I’m Kanishk 👋

MSc Artificial Intelligence candidate building production-grade AI systems, real-time data platforms, and distributed backends that (usually) behave in production.

• 🧠 Research + engineering combo: 3 publications, hands-on AI internships, and too many tabs open.
• 🏆 1st Place — QuantiHack 2026 (Team APEX, 800+ participants).
• ⚙️ Focus: Python, ML systems, event-driven architecture, real-time pipelines, cloud deployment.

---

🔥 Featured Project (Top Pick)

The Pre-Mortem Machine

Forensic supply-chain intelligence system built at QuantiHack 2026.

• Predicts CPG company stress fractures before market pricing (avg 19-day lead time).
• Multi-agent style architecture over 6 live public data streams.
• FastAPI + React system with composite scoring, canary ranking, and blame-chain analysis.
• Winner project: £3,750 + £5,000 Anthropic credits.

We run the autopsy before the patient dies. (Yes, that’s the actual product vibe.)

---

🚀 Other Strong Projects

• Multi-Strategy Backtesting Engine
  Professional quantitative trading framework with CCXT/yfinance pipelines, HRP + Black-Litterman optimization, VaR/Greeks, and RL agents (PPO/SAC).

• Enterprise Data Analytics & BI Platform
  End-to-end analytics stack with Airflow + Kafka + ClickHouse + FastAPI + MLflow, including dimensional modeling and sub-100ms query targets.

• AI-Powered Event-Driven Order Processing System
  Microservices + Saga orchestration + RabbitMQ + Transactional Outbox, with fraud detection and anomaly monitoring.

• AI Finance Analyst
  Next.js + Plaid + Claude-powered personal finance assistant for automated savings insights and secure analytics workflows.

• AI Cost Profiler
  Lightweight SDK for tracking LLM costs and usage in agent workflows.

---

🛠️ Core Stack

Python TypeScript C++ FastAPI React Node.js TensorFlow
Docker Kubernetes AWS Kafka RabbitMQ MongoDB Redux

---

🧪 Experience Highlights

• FLSmidth (AI Engineer Intern): Led 3-person team to ship a 3D reconstruction workflow; reduced inspection operational complexity by 40%.
• AIcade (AI Engineer Intern): Improved asset generation workflow performance by 24% and supported scalable AI character generation pipelines.

---

📚 Publications

• 3D Surface Reconstruction research (IEEE, 2025)
• IoT-Based Proximity Detection for Object Tracking (IEEE ISES, 2024)
• Deep Learning for Early Detection of Neurodegenerative Disorders (arXiv preprint)

---

🌐 Connect

• GitHub: @koopatroopa787
• Email: yashk242810@gmail.com