Bhautik Vekariya

I’m an aspiring Data Scientist and Machine Learning Engineer specializing in end-to-end ML pipelines, deep learning models, and MLOps systems. I enjoy turning messy data into reliable predictive solutions, and I’m particularly motivated by how well-designed pipelines translate into business value through dependable model deployment.

In my Data Analytics Intern role (Jun 2024 – Aug 2024) through IBM SkillsBuild and CSRBOX Foundation, I applied statistical hypothesis testing (t-tests, ANOVA) to validate feature engineering and improved model accuracy by 15%. I also designed 5+ interactive dashboards using Power BI and Tableau with real-time filters and drill-down capabilities, helping stakeholders make clearer, data-driven decisions through strong visual storytelling.

To deepen my applied expertise, I completed a project-based Data Science Mentorship Program (May 2025 – Jan 2026) at CampusX, covering a 10-module curriculum across Python, statistics, machine learning, deep learning, and MLOps fundamentals. I built real-world case studies using EDA, feature engineering, and predictive modeling, reinforcing the practical habits needed to ship models—not just prototype them.

My projects reflect how I build production-minded systems. In my AQI Prediction MLOps Pipeline, I created a production-grade REST API for real-time AQI predictions across 29 Indian cities using XGBoost and FastAPI, deployed with Docker and an automated CI/CD approach; I reduced the model’s memory footprint by 70% while cutting batch response time from 30s to under 3s, and I used MLflow + DVC for experiment tracking/version control with zero deployment failures over 4 months. In my AI Interview System, I built an end-to-end AI mock-interview platform with resume parsing, LLM-powered role-aware question generation, and voice-enabled flows (TTS/ASR), backed by Cassandra and a ChromaDB vector store, then deployed a microservices stack on Kubernetes (EKS) with CI/CD via GitHub Actions, Argo CD GitOps, and observability using Prometheus/Grafana and Jaeger alongside Kafka event streaming.