# Prakarsh Goel **Role:** Data Science · QA · AI/ML **Location:** Electronic City, Bengaluru **Status:** QA Intern at Evaratus (formerly Scaler AI Labs) ## Current focus - QA at Evaratus — getting better at testing, debugging and finding edge cases in RL environments. - Building AI/ML projects: a DQN agent that plays FlappyBird and a Transformer–RAG system that catches SMS scams. - Two research papers under review: AI-based scam detection and groundwater time-series forecasting. ## Contact - Email: prakarshgoel1575@gmail.com - WhatsApp: +91 63630 40595 ## Index - [Work](https://prakarsh.tech/work) - [Projects](https://prakarsh.tech/projects) - [Paper shelf](https://prakarsh.tech/paper-shelf) - [Certifications](https://prakarsh.tech/certifications) - [LeetCode](https://leetcode.com/u/pg1575/) - [Blog](https://prakarsh.tech/blog) - [Resume](https://prakarsh.tech/resume)
hola · i’m ◢
Data Science · QA · AI/ML
Prakarsh
Goel.
I enjoy breaking software, building projects, and exploring how AI can solve real-world problems.
Mostly in Python. Reinforcement learning rabbit holes are my version of touching grass.
Currently teaching agents to play games (and losing to them)
Final-year B.Tech Data Science at
QA Intern at EvaratusElectronic City, Bengaluru
Current focus
- QA at Evaratus — getting better at testing, debugging and finding edge cases in RL environments.
- Building AI/ML projects: a DQN agent that plays FlappyBird and a Transformer–RAG system that catches SMS scams.
- Two research papers under review: AI-based scam detection and groundwater time-series forecasting.
§01 · status
Where, what, who.
📍 Electronic City, BengaluruLive · esri
On loop
Now playing
Resonance
HOME
I put “Data Science” and “QA” on everything because I like building models and then trying very hard to break them.
§02 · stack
What I reach for.
A snapshot of my employability as of this morning. Tomorrow’s snapshot is still training.
Data science5
Turning spreadsheets into opinions, then checking them.
- Data analysis
- Statistics
- Data visualisation
- Predictive modelling
- Time-series forecasting
Machine learning7
From gradient boosting to agents that learn by failing.
- Reinforcement learning
- DQN
- Transformers
- RAG
- LSTM
- XGBoost
- Neural networks
QA & testing3
Finding the edge case before production does.
- RL task validation
- Edge-case analysis
- Debugging
Languages3
Pick the right tool for the layer.
- Python
- SQL
- C++
Libraries & tools7
What’s open in the other tab.
- PyTorch
- NumPy
- Pygame
- Matplotlib
- SpeechRecognition
- pyttsx3
- Git
§03 · git activity
605 contributions this year
§04 · coding profile
300 problems solved - on LeetCode
§05 · credentials