Trishala Thakur
Machine Learning Data Scientist @Harris School of Public Policy at the University of Chicago
Signup · Get unlimited contacts
WORK HISTORY
Machine Learning Data Scientist @Harris School of Public Policy at the University of Chicago
Chicago, IL, US
Built agent-orchestrated pipeline (Gemini 2.0 Flash) to automate face-name matching at 92% F1 score, eliminating manual annotation Leveraged few-shot prompting to classify page layouts and segment a 1TB+ yearbook corpus by complexity with 71% accuracy Prompted GPT-5-nano to extract insights from 1.2M papers, enabling computational social science analysis and saving 12k review hours
EDUCATION
JSS SCIENCE AND TECHNOLOGY UNIVERSITY
Bachelor of Engineering - BE
University of Colorado Boulder
Master of Science
ABOUT TRISHALA THAKUR
I’m a data scientist who has spent the past ~5 years solving problems across diverse domains: energy infrastructure, environmental science, and social science.My journey started in 2021 at Schneider Electric, where I worked on predictive maintenance for industrial equipment. I focused on previously untapped mechanical failure signals from sensor data, building models that helped engineers anticipate failures, reduce downtime, and make better design decisions.During my master’s, I transitioned from classical machine learning to deep learning, working on environmental challenges like drought prediction. I designed a convolutional neural network to identify drought patterns in the western U.S. using satellite-derived data, exploring how representation learning can uncover subtle signals in spatiotemporal datasets.Currently, I work at the intersection of AI and social science, building large-scale pipelines that extract structured insights from massive image and text datasets. This includes designing agent-based systems, LLMs for information extraction, and developing computer vision pipelines to enable research at scale.Across all of this, what I’ve enjoyed most is working across domains, learning from experts and building systems that make their work faster, more scalable, and more impactful.Languages & Tools: Python, SQL, PyTorch, TensorFlow, scikit-learn, Hugging FaceCore Areas: Machine Learning, Deep Learning, LLMs, RAG, Computer Vision, NLPSystems & Infra: GCP, Docker, APIs, Data Pipelines, Model DeploymentMethods: XGBoost, LightGBM, Feature Engineering, OpenCVLet’s connect - t••••••@gmail.comI’m always excited to collaborate on interesting problems or explore opportunities in data science and applied AI
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.