Guest: Mona Rakibe (LinkedIn)
Company: Telmai (Twitter)
Show: Let’s Talk About AI

The increased velocity and scale of data is making it challenging to predict the health of the data, particularly when enterprises are making key business decisions based on the data. Data observability company, Telmai, aims to make it easier for organizations to understand their data and to investigate anomalies.

In this episode of TFiR: Let’s Talk About AI, Mona Rakibe, Co-Founder and CEO at Telmai, talks about the company and how it is helping companies improve their data quality and investigate anomalies. They go on to talk about the company’s journey so far, some of the key capabilities of the platform, and what sets them apart from competitors.

Key highlights from this video interview:

• Rakibe talks about the evolution of enterprise data strategy over the years and how although now there is a focus on real-time analytics and AI/ML-based initiatives to handle the increased velocity and scale of the data, we are still using legacy systems to predict the health of the data. Rakibe introduces us to the data observability company, Telmai, and talks about the origin of the name and the problem they are trying to solve.
• Telmai was created to observe the data itself that is fueling the key business decisions and the AI/ML initiatives in order to predict the health of the data. Rakibe explains what sets Telmai apart from the crowd: supporting different data sources in different formats, doing ML-based anomaly detection, and doing it without adding any latency to the data.
• The solution is agnostic to a vertical but Rakibe believes that verticals like financial services and healthcare, and industries that have reached a certain level of maturity with their data initiatives, are quicker to identify the problem of scale and understand the solution differentiation.
• Rakibe talks about the role of AI in the company: seeing where it can be used internally to see which features can be used or developed using generative AI faster, cheaper, and better. She discusses the use cases where they are using generative AI.
• Rakibe tells us about research Telmai did recently that there was a 10-15% precision difference if you pre-process the data a little and control the noise level of the data that goes through model training. She talks about the cost implications of having to retrain models if you push data and start training without any preparation work.
• Telmai provides a full end-to-end solution around data reliability with a data observability foundation, but also to help organizations learn about their data and establish some data quality guardrails around data quality, understand anomalies in the data, and understand the parametric monitoring system when migrating data.
• Rakibe discusses how they had realized that the biggest business impact was going to be with enterprises and as such they had to change their product direction accordingly. In September this year, Telmai announced its most significant release to the market to date, featuring seven category-defining features designed to accelerate data observability adoption for the enterprise.

This summary was written by Emily Nicholls.