kanish = {
"name" : "Kanish Tyagi",
"role" : "Data Scientist & AI/ML Engineer",
"education" : "M.S. Data Science @ UT Arlington (Dec 2026)",
"background" : "3+ yrs QA Engineer → now building AI systems",
"location" : "Arlington, TX 🇺🇸",
"open_to" : ["Data Scientist", "Data Analyst", "ML Engineer"],
"currently" : "Building Adaptive AI Tutor with RL + Claude API",
"superpower" : "Turning messy data into intelligent decisions 🧠",
}- 🔭 Working on: Adaptive AI Study Tutor — UCB1 Multi-Armed Bandit RL + Claude API for real-time personalized learning
- 👯 Collaborating on: LLM-powered apps, RAG pipelines, NLP & predictive analytics with real-world impact
- 🤝 Looking for: My first Data Scientist / Analyst role — open to referrals, networking & mock interviews!
- 🌱 Learning: MLOps, LLM fine-tuning, AWS/GCP deployment — bridging notebooks → production
- 💬 Ask me about: ML pipelines, LangChain/RAG, Python, RL systems, or my QA → DS journey
- ⚡ Fun fact: Cut post-release defects by 30% as a QA Engineer — now I build AI that teaches itself. Debugging models > debugging code 🚀
| 🎤 Speaker | 🏆 Hackathon | ✅ Certified | 📜 Certified |
|---|---|---|---|
| UTA Business Analytics Symposium 2026 | 2nd Runner-Up — AWS GenAI Hackathon | ISTQB Foundation Level | Google Data Analytics Professional |
| Project | Stack | Highlights |
|---|---|---|
| 🤖 Adaptive AI Study Tutor | Python · UCB1 RL · Claude API · Streamlit · SQLite | 18 config arms · Real-time adaptation · Full-stack |
| 🏥 Health Center Attrition Predictor | Scikit-learn · XGBoost · Pandas · GridSearchCV | 87% accuracy · 1,400+ records · 3 models compared |
| 🏠 Airbnb Price Insights | Pandas · Matplotlib · Streamlit · EDA | 50K+ listings · Interactive dashboard · Geo pricing |
| 👥 Customer Segmentation | K-Means · PCA · Scikit-learn · Seaborn | 4 segments · PCA dim reduction · Marketing insights |
| 📈 Market Data Forecasting | ARIMA · Pandas · Matplotlib · Time Series | Automated pipeline · Exponential smoothing |
