Data Scientist | GPU-Accelerated MLOps Practitioner | End-to-End ML Pipeline Developer | Real-Time Forecasting Specialist
- Email: thejaskiran99@gmail.com
- LinkedIn: linkedin.com/in/thejasbk
- GitHub: github.com/ThejasBK
- Medium: medium.com/thejaskiran99
- Credly credly.com/thejas-kiran
👋 Hi there! I'm Thejas Balenahalli Kiran, a results-driven Data Scientist with experience in designing and implementing end-to-end pipelines for real-time sales forecasts and optimizing forecasting service pipelines for significant performance improvements. I specialize in leveraging GPU multiprocessing, reducing cloud costs, and developing impactful Python packages. I am passionate about applying machine learning and deep learning techniques to solve complex business problems and improve product recommendations and operational efficiency.
- Google Cloud Platform (GCP)- Vertex AI, BigQuery, Feature Store, Dataflow, Dataproc, Cloud Storage
- Machine Learning frameworks - PyTorch (GPU acceleration), TensorFlow
- MLOp & Pipeline Orchestration - Kubeflow, Docker, Kubernetes, Airflow, Git (version control)
- Programming Languages - Python, SQL, R
- Big data & data processing - PySpark, Databricks, Apache Kafka
- Web Development Frameworks - Flask, Django
Aug 2023 – Present
- Led a team in designing and implementing an end-to-end pipeline for real-time sales forecasts.
- Optimized a forecasting service pipeline using GPU multiprocessing, reducing processing time by 40% without increasing model run costs.
- Reduced BigQuery cloud costs by 95%, improving query speed by 70%.
- Developed and contributed to a Python package adopted across multiple teams within Sam's Club.
- Collaborated with a cohort from the University of Arkansas on graph neural network for item similarity, improving product recommendations.
- Participated in hackathons focused on LLMs and item similarity with deep learning, achieving top results.
- Technologies Used: Python, SQL, GCP (BigQuery & BigQuery ML, Dataproc, Cloud Storage, Vertex AI), PyTorch, PySpark, Docker, Kubernetes, Airflow, Kubeflow, Git, Databricks, Apache Kafka
Dec 2021 – Apr 2024
- Improved image embedding algorithm performance by 25% with GPU acceleration.
- Implemented a topic modeling algorithm for image categorization.
- Created and deployed a survey portal with JavaScript, incorporating active learning for user preference capture.
- Enhanced consumer marketing strategies by optimizing image mining algorithms.
- Technologies Used: Python, TensorFlow, GPU computing, Django
May 2023 – Aug 2023
- Developed dashboard to track key customer metrics and business KPIs, reducing reporting time by 60%.
- Optimized an end-to-end data pipeline, improving processing and storing speed by 25%.
- Integrated diverse datasets via web scraping and APIs for market research and decision-making.
- Technologies Used: SQL, Python
Jun 2022 – Aug 2022
- Implemented a deep learning algorithm, boosting sales forecasting accuracy by 20%.
- Reduced computing costs by 94% and model runtime by 86% by optimizing the forecasting pipeline.
- Built Spark-based pipeline for larger datasets enabling faster model training and data analysis.
- Technologies Used: Python, PySpark, PyTorch, GCP (BigQuery, Dataproc, Vertex AI), Databricks
British Airways Data Science Job Simulation on Forage | [Link to your Forage Project if available, or a GitHub repo if you created one for it]
Jan 2024
- Applied data analysis techniques to optimize business operations and enhance decision-making.
- Built an interactive dashboard to track customer interactions, improving insights into KPIs.
- Designed and implemented data pipelines for seamless data integration and transfer across systems.
- Technologies Used: Python, SQL
University of Colorado Boulder, USA | 2021-2023
- GPA: 3.99/4.0
VTU, India | 2016-2020
- GPA: 8.38/10.0
- Unlocking the potential of CNNs in time series forecasting
- Unleashing the Power of GPUs: A Deep Dive into Their Role in Accelerating Deep Learning
- Teaching AI to each art: A GenAI assistant for painters
- Linear regression using PyTorch
- Analysis, Visualization, and Prediction of COVID-19 Pandemic Spread Using Machine Learning (Springer)
- Wheelchair Control and Home Automation using hand gestures (European Alliance for Innovation)
- Home Automation and Wheelchair Control using Air Gesture: A Survey (PICES)
Last Updated: May 2025