Tavro Data ROI Agent

Sunil Soares, Co-Founder & CEO, Tavro

Assessing the ROI of Data is an Extremely Cumbersome Task Because Data Leaders are Not Financial Analysts
I spent a lot of time with my other company (YDC) on Data Economics including the computation of the return on investment on data projects. The computation of Data ROI tends to be extremely cumbersome because it entails the intersection of two different skills: Data Literacy and Financial Savvy.

The typical steps in the computation of Data ROI are as follows:

  1. Identify the use case with internal stakeholders
  2. Gather additional information on the use case
  3. Agree on overall use case objectives in terms of cost reduction, revenue improvement, or risk mitigation
  4. Research business drivers based on industry metrics
  5. Tweak business drivers based on company-specific data
  6. Populate financial values for business drivers based on research and interaction with internal stakeholders
  7. Iterate and align with internal stakeholders


Tavro’s Data ROI AI Agent

We used a simple prompt (company name has been redacted), 

“Generate a business case for AI-driven order entry automation for [company].”

The agent implemented a number of tasks and produced a business case in CSV format as shown in the figure above. The business case included fields for Driver Name, Driver Description, Driver Type, Driver Comments, Driver Value, Driver Formula, and Sources. The company-specific information has been redacted.

Assessing the Results of the Agent and Human-in-the-Loop
Admittedly, the first pass of the business case appears to yield minimal value of $3.75 million. From prior experience, I know that just producing a first draft of the business case is likely a big lift for a data leader and might take several days. The Tavro Data ROI Agent is able to automate this process down to minutes.

Tavro’s Data ROI Agent automated steps 2-6 in the Data ROI process mentioned above. However, we still need humans to do the following:

  1. Identify use cases
  2. Align with stakeholders to tweak the business case drivers and values to produce more realistic results