Vishwas Prabhakara
Machine Learning Engineer @ IISc | Forecasting & Applied AI for Environmental Systems | Python, Time Series, GCP
Greater Bengaluru Area
2+ years of work experience
Oct 20251 yr 1 moPresent
• Lead ML forecasting work for Bengaluru groundwater planning, building ward-level groundwater prediction pipelines across 100+ wards with forecast horizons up to 6 months.
• Improved model performance through feature engineering, lag analysis, climate and urban variable integration, and two-stage feature selection using Pearson correlation and mutual information; achieved ward-level R² up to 0.9 in validation.
• Built end-to-end ML workflows in Python: data preprocessing, feature extraction, model training, evaluation, reproducibility checks, and visualization for research and stakeholder-facing analysis.
• Translated city-government requirements from the IISc × BWSSB collaboration into practical ML deliverables for operational groundwater decision support.
• Presented AI/ML groundwater research at IISc Open Day 2025 to visitors from academic, government, and public-sector backgrounds.
Apr 20262 mosMay 2026
India
Selected for Google Cloud Gen AI Academy APAC 2026, Cohort 1, focused on building production-oriented GenAI applications with Gemini, Google Cloud, and cloud-native deployment workflows.
Built Sutra, a six-agent AI chief-of-staff system orchestrated through Gemini function calling and deployed on Google Cloud Run.
• Designed multi-agent workflows with structured tool orchestration and human-in-the-loop confirmation
• Built a FastAPI backend with SQLite persistence and a React + TypeScript frontend
• Implemented real-time execution traces using Server-Sent Events for transparent agent behavior
• Managed cloud deployment, architecture documentation, project demo, and technical pitch
Oct 20241 yrSep 2025
• Built the first version of the ward-wise groundwater forecasting pipeline at IISc CST, combining hydrology, climate, and geospatial datasets for Bengaluru groundwater-level prediction.
• Prototyped ensemble ML models, engineered time-series features, and evaluated model behavior across ward-level variability and forecast horizons.
• Contributed to a computer-vision pipeline for microplastic detection from µ-FTIR imagery in an IISc–University of Cologne collaboration, using classical CV methods for particle and fiber extraction.
• Worked with PhD scholars and senior researchers on research-grade ML workflows, data validation, and scientific reporting.
Mar 20227 mosSep 2022
Bengaluru
• Worked on Lip-to-Speech synthesis pipelines involving audio-visual speech reconstruction, model optimization, and inference workflow improvements.
• Built and trained machine learning models in Python and TensorFlow for client-specific project requirements.
• Assisted with model integration and deployment workflows, including testing, debugging, and performance validation.
• Collaborated with project teams to convert experimental ML workflows into usable application components.
Vishwas’s AI Stack
Education
Master of Engineering - MEng, Artificial Intelligence
Manipal School of Information Sciences
20232 yrs2025
Bachelor of Engineering - BE, Computer Science
K S school of engineering and management
20194 yrs2023
Join Vishwas and 10k+ other professionals on
Tal is the app where bosses meet engineers & builders directly, without middlemen
Download Tal Boss