Vaibhav Vishwakarma

I’m an AI & ML Intern with a B. Tech in Computer Science (Data Science) and an AWS Certified Solutions Architect – Associate (SAA-C03). I bring a product mindset to transforming messy inputs into reliable, decision-ready data.

At Devmani Traders, I built a production Python ETL pipeline processing 80,000+ mutual fund records monthly across variable Excel schemas. I used word embedding-based semantic column matching to handle schema drift, deployed the pipeline on GCP Compute Engine with daily automation, and implemented logging, validators, and SQLAlchemy sync to maintain high data integrity. I also engineered a historical aggregation system pulling 10+ years of mutual fund scheme data from AMFI, automating daily metric computations and backend synchronization—eliminating a multi-day manual workflow to minutes.

My projects reflect how I like to engineer end-to-end solutions: a real-time streaming “Machine Anomaly Alarm System” using a two-stage decision logic and an adapted SW-WaveNet model (0.75 pAUC vs 0.72 baseline), plus a containerized Kafka + Spark Streaming microservices pipeline for industrial water pump monitoring. I’ve also built a FastAPI + Vanilla JavaScript app (“Am I Debarred”) with linked-list based server-side logic and a decoupled Render/GitHub Pages deployment, focusing on clean iterations and practical adoption.