← Back to Projects
This project is a sub-project of AgCloud
Sensors
Mentored by: Vast Data
Sensors - Cloud-based platform for agricultural data management and analytics
React
Node.js
PostgreSQL
AWS
Docker
Kubernetes
Microservices
Description
Sensors sub-project of AgCloud. A comprehensive cloud platform for managing agricultural operations, data, and analytics. Provides centralized storage, processing, and visualization of farm data including crop monitoring, weather integration, equipment management, and predictive analytics. Features include multi-tenant architecture, real-time dashboards, and API integrations.
Team Members
Cohort: Data Science Bootcamp 2025 (Data)
Responsibilities:
Set up and configured MinIO (S3) for data storage with a Lifecycle Policy.
• Built an anomaly detection model using scikit-learn and integrated it into a real-time Apache Flink streaming pipeline.
Connected the system to Kafka and MQTT pipelines and integrated PostgreSQL for data storage and processing.
Deployed a complete Edge Device including a visual GUI monitoring interface.
Integrated Grafana for real-time visualization and system monitoring.
Generated embeddings and built a pgvector-based similarity search API.
• Packaged and ran the entire system using a unified Docker Compose deployment
...and more contributions not listed here
Responsibilities:
Developed a real-time adaptive Sensor Modeling Engine with dynamic baselines, bias correction, seasonal and diurnal pattern learning, and early anomaly detection—significantly improving signal quality and reducing false alerts.
Architected a scalable, high-throughput ingestion pipeline for S3/MinIO using Python, adaptive flow control, and optimized multipart uploads, achieving over 17× throughput improvement.
Implemented a production-grade Flink Writer supporting atomic, low-latency, high-volume writes to PostgreSQL with isolated workers, intelligent retry/timeout logic, and controlled backpressure, ensuring stable performance during database slowdowns.
Built a secure FastAPI-based authentication API implementing JWT with token rotation, hashed service accounts, and strict access control policies.
Implemented automated zone-based analytics to compute real-time statistical metrics (mean, median, min/max, standard deviation, anomaly counts) across field regions. Developed a GPS-driven heatmap engine that interpolates sensor data to visualize environmental conditions ( e.g. soil moisture, temperature, humidity), enabling fast and informed decision-making
...and more contributions not listed here
Responsibilities:
Deployed a fully compatible MINIO (AWS S3) object-storage server within a Dockerized environment, enabling scalable and consistent storage operations for internal data services.
Built a real-time sensor integrity and anomaly-detection pipeline using Apache Flink, integrated with Apache Kafka to process high-throughput streaming data and provide continuous monitoring and operational insights.
Developed a generic, schema-aware API microservice in Python, providing secure CRUD access over authorized database tables, with support for dynamic filtering and reusable data-access logic.
Developed a dynamic GUI dashboard for real-time sensor visualization using PyQt6 and pyqtgraph, featuring interactive charts, auto-refresh, and structured alert reporting.