wellsight-ai

• 

  wellsight-ai-architecture

Inspiration

The oil and gas industry loses billions each year to unexpected equipment failures and unplanned downtime, largely because maintenance is still reactive—fixing problems only after they occur. While analyzing public well data from the Oklahoma Corporation Commission covering over 125,000 wells, we identified a clear opportunity to apply modern AI and real-time analytics to shift this model. The industry already generates vast amounts of sensor data, yet much of it remains underused, despite the fact that failures are often preceded by detectable warning signs in pressure, temperature, vibration, and flow readings.

Our inspiration crystallized around a simple but powerful idea: if a single day of downtime can cost operators $50,000–$100,000, why wait for failure at all? By recognizing that anomalies typically appear days or weeks in advance, we envisioned a proactive, data-driven platform that could predict failures before they happen. This question—“What if equipment failures could be predicted 30 days in advance?”—led to the creation of WellSight AI, designed to help operators prevent downtime, reduce costs, and fundamentally change how maintenance is approached in oil and gas operations.

Built With

• docker
• fastapi
• influx
• minio
• mqtt
• posgre
• python
• qdant
• rabbitmq
• raindrop
• vultr