DESCRIPTION
In this episode, Matt Oberdorfer, CEO, EOT shares insights with Host, Elena Melchert, on how AI and time series models are transforming upstream oil and gas operations, emphasizing practical applications, data cleaning, operator training, and the future of AI-driven decision making.
LINKS
- eot.ai : our software company which is focused on time series AI models that can be used by operators and engineers (without coding)
- chronx.ai : the actual software that is using a time series transformer model to enables future discovery dashboards where engineers can train, learn and validate their private AI models for equipment they are responsible for.
- A scientific summary on time series transformer model in AI: https://www.sciencedirect.com/science/article/pii/S1574013725001595 :
- How a transformer model works (very technical) https://www.alignmentforum.org/posts/pHPmMGEMYefk9jLeh/llm-basics-embedding-spaces-transformer-token-vectors-are
- Here is a prompt / question that anyone can put into ChatGPT / Claude / etc. I like this prompt a lot because it will work in anyones context (try it out for yourself and copy and paste it into AI).
“Explain time-series transformer models in simple terms for an oil & gas upstream engineer. Explain how they learn from large multivariate histories of sensor data (pressure, temperature, vibration, flow, etc.) and discover complex temporal relationships to understand equipment behavior, answer questions about the past, and predict what happens next. Use practical examples with pumps, compressors, ESPs, wells, and production facilities. Compare transformers with traditional AI/ML and explain why and when transformers can be better, particularly when dealing with large amounts of historical data, many interacting signals, and complex relationships across time.”
Brought to you on OGGN, the largest and most listened-to podcast network for the oil and energy industry.
Sign up for OGGN newsletters here.
Visit the OGGN Merch store here.
Curious about podcast advertising? Learn more here.
More from OGGN …
Podcasts
LinkedIn Group
LinkedIn Company Page
Podcast: Play in new window | Download