Biplov P.
Data & AI Engineer with 4+ years of experience in designing and implementing cloud-native data pipelines, scalable machine learning systems, and production-grade APIs. Proficient in Spark-based ETL, cloud infrastructure, and AI-driven solutions, including Retrieval-Augmented Generation (RAG) and computer vision. Demonstrated success in delivering high-quality data products across e-commerce, healthcare, and financial domains for global clients.
- Managed and optimized end-to-end Google Cloud Platform (GCP) infrastructure, including BigQuery, GKE, GCS, Vertex AI, Cloud Functions, and Cloud Run.
- Designed, developed, and maintained scalable ETL pipelines, integrating data from 8+ platforms (Zoho, Aroflo, SEMrush, Google Ads, GA4, Google My Business, Pronto, Pipedrive) into BigQuery.
- Built and deployed RESTful APIs and FastAPI-based microservices, enabling seamless data processing and system integration.
- Automated workflows using Make (Integromat), Zapier, and n8n, improving operational efficiency and reducing manual intervention.
- Developed RAG-based AI solutions for duplicate detection and intelligent product recommendations.
- Created and maintained business intelligence dashboards, delivering actionable insights for data-driven decision-making.
Designed and deployed RAG-based AI solutions, integrating them into web applications to enhance search and decision-making capabilities.
Developed and optimized computer vision models for waste classification, improving accuracy and model performance.
Generated high-quality synthetic datasets using TimeGAN, SDV, and Diffusers to support model training and data augmentation.
Leveraged tools and frameworks including Python, Apache Spark, GCP, MLflow, Vertex AI, and Hugging Face Transformers for scalable AI/ML development.
- Developed data acquisition pipelines using Python and C++ for efficient sensor data collection and processing.
- Designed and implemented long-range (up to 10 km) data transmission systems with real-time visualization.
- Built and deployed Flask-based analytics applications on Raspberry Pi (Linux/Debian) for edge computing.
- Optimized system performance and reliability across hardware-software integrated environments.
- Delivered 18+ international projects across machine learning, data engineering, and cloud automation.
- Processed and analyzed 100GB+ datasets using PySpark and XGBoost for scalable model training.
- Developed and deployed Dockerized Python applications and FastAPI-based services.
- Built computer vision prototypes for tumor detection, contributing to research-oriented solutions.