I build reliable, production-ready software at the edge of applied AI and large-scale data.

• Implement a stereological ore estimation module and a multimodal AI-driven EDA engine that automates pattern discovery, anomaly flagging, and volume/grade estimation across geological, assay, and survey datasets — replacing a manual interpretation workflow that took geologists 2–3 hours per site survey.
• Develop a high-performance point cloud and geological data visualization system on Potree and 3D Tiles, rendering large-scale subsurface LiDAR and terrain scans directly in-browser for real-time interactive interpretation.
• Integrate Copernicus multispectral satellite data into the core ingestion pipeline to power bauxite deposit prediction over user-selected bounding boxes, extending the platform beyond borehole-only datasets.
• Design and ship a full CI/CD pipeline on GitHub Actions that promotes releases across staging and production with blue-green deployment, delivering zero-downtime cutovers and instant rollbacks.
• Harden production infrastructure across 4 core subsystems with Docker container health checks, database connection pooling, and automated nightly backups — cutting idle compute cost by 70%.

• Optimized the database layer with 3 migrations indexing 5 hot tables, eliminating full scans of 470K-row and 137K-row tables to cut page loads from 860ms to under 100ms — a 10x speedup across every admin page.
• Rearchitected dashboard and reporting endpoints that shared an N+1 query loop, folding hundreds of per-equipment round trips into a single index-backed windowed SQL query and cutting median response times from 78s to below 1s.
• Offloaded synchronous Python ML report generation (ferrography classification and predictive wear modelling) onto 2 dedicated database-backed queue workers, taking report pages from 47–120s to near-instant.
• Containerized the full production stack with Docker Compose across 4 services — Laravel 9/PHP-FPM, Caddy with automatic HTTPS, a Flask ML inference API, and queue workers — isolating ML queries from public traffic.
• Built an automated end-to-end ML retraining pipeline for the ferrography classifier, versioning training datasets on Cloudflare R2 so the model improves as new labelled lab samples accumulate in production.

• Designed the platform's UI/UX in Figma from scratch and built the client-facing React.js frontend, a Node.js/Express API server, and a Python FastAPI microservice for ML and geological computation.
• Built a borehole desurvey and compositing pipeline from scratch, converting raw Collar/Survey/Assay CSVs into 3D spatial trajectories across 39 collars, with length-weighted assay compositing and lode calculations to delineate mineralized zones.
• Engineered a fault-tolerant, chunked ingestion pipeline streaming 100+ GB of LiDAR, tabular, and geospatial data with error recovery, schema validation, and resumable chunk retries.
• Architected and deployed the full-stack production infrastructure on AWS EC2 with Nginx reverse proxy, Docker containerization, Postgres, and AWS S3 for point cloud and satellite imagery storage.

Winner, Smart India Hackathon 2024

Led a 5-member team building an AI-driven water resource optimization platform at a national-level innovation competition.

Winner, Nebius Everyday AI Hackathon 2025

Hosted by Nebius & Sprint.dev, recognized for applied AI product development.

6+ verified certifications

Credentials from IBM and Microsoft Azure spanning cloud computing, AI/ML, and data fundamentals.