A Reinforcement Loop
Reward signal: the logic of a checkmate, the resolve of a power chord.
Hey there! I’m Vikas Ravikumar Karjigi, a curious and driven Data Scientist & AI/ML Engineer with 3+ years of hands-on experience turning ideas into real-world impact. Right now, I’m leveling up my skills through a Master’s in Data Science at Illinois Institute of Technology, Chicago, where I proudly hold a 4.0 GPA.
I spent over two years at Boeing, building scalable pipelines and automation systems that made data work smarter. More recently, I’ve been diving into cutting-edge research on secure and decentralized LLMs. Along the way, I’ve built an AI sports highlight generator with an 88% F1 score and slashed processing time by 98% in a real-time fraud detection pipeline. I love crafting systems that are fast, smart, and actually useful.
Lately, I’ve been especially excited about the rise of generative AI, LLMOps, and all things AI infrastructure. When I’m not building with code, you’ll probably find me deep in a game of chess or strumming away on my guitar. Always exploring, always creating.
My Tech Arsenal
From code to cloud: Here’s what powers my work in building elegant, intelligent solutions.
Career Highlights
A track record of turning data-driven insights into measurable results.
Jun 2025 – Present
- Designed a reinforcement learning pipeline using PPO in OpenAI Gymnasium (Ant-v4) with MuJoCo physics engine and Stable-Baselines3.
- Integrated trained agents into a custom Unity 3D environment using socket programming and Oculus VR for immersive RL visualization.
Jun 2025 – Present
- Mentored graduate students in data wrangling, feature engineering, dimensionality reduction, and model evaluation using Python, R, SQL, and Pandas.
- Reinforced data science concepts during lab sessions, project reviews, and code debugging.
Sep 2024 – May 2025
- Built a computer vision pipeline to detect and classify highlight-worthy moments in volleyball using OpenCV and motion-based features, achieving 88% F1-score.
- Automated 1–2 minute highlight reel generation, reducing manual editing time by 80%.
Aug 2022 – Jul 2024
- Improved booking efficiency by 30% through SQL pipelines across MySQL and MSSQL for real-time KPI dashboards.
- Saved $60K+ annually by reducing reporting latency by 98% using stored procedures and indexing.
- Built standardized SQL templates and automated 500+ data validation checks with Python to ensure compliance.
Jul 2019 – Sep 2019
- Built a Python-based ETL pipeline using Pandas and MySQL to clean and normalize 25K+ records.
- Reduced manual preparation by 60% and enabled faster feature extraction for ML.
Passion Projects
Projects where the purpose, not just the technology, is the primary driver.
Education
Learning never stops, but here’s where it began.
Core Courses: Big Data Technologies, Data Preparation and Analysis, Regression, Machine Learning, Time Series, Monte Carlo Methods, Project Management.
Completed comprehensive coursework in electronics, communication systems, and signal processing.
Certifications
A commitment to continuous learning and professional growth.