TIS helped BMS treasury standardize and automate forecasting. The treasurer wants AI to make the tool even better.
Bristol Myers Squibb treasurer Sandra Ramos-Alves loves automation and is a self-described “strong believer” in tapping the expertise of fintechs whose solutions can help BMS transform treasury. TIS is one of those fintechs and its CashOptix tool—which includes cash forecasting capabilities—has enabled BMS to produce an aligned view of cash flow from one source of truth used by everyone across the globe.
- Ms. Ramos-Alves says the TIS implementation is a fundamental part of a broader transformation that began in 2021 focusing on people, processes and technology during which the BMS treasury team has embraced automation and tackled more than 50 projects.
- “This is just one of the many things that we’re working on in terms of the evolution of treasury and building a treasury of the future—TIS is just part of the solution for us,” she said at a recent session of NeuGroup for Life Sciences Treasurers.
Eager for AI. The BMS cash forecasting and liquidity management journey—described in a detailed deck presented during the session and available to members upon request—is not complete. Ms. Ramos-Alves is eager for TIS to roll out improvements built with AI as treasury focuses on one of its key objectives: working capital management. “We launched our cash leadership office last year so we’re working really hard to optimize our working capital,” she said.
- “We’re working to see when we take this tool to the next step and how we’re going to really use the working capital module,” she added. “We’re excited to see how we can take cash forecasting to the next level with the addition of [generative] AI.”
- The good news is that TIS is working closely with its customers to identify ways that AI and ML can bring value to cash forecasting and working capital management processes, according to Chief Product Officer Jon Paquette.
- He said the company recently held a customer workshop to discuss various AI application concepts for treasury. These include both AI enhancements to forecasting logic as well as simplifying the overall platform configuration through the use of generative AI chatbots.
The data challenge. Tanya Lurye, associate director of US Cash Management at BMS, is looking forward to the arrival of the TIS AI chatbot because, in part, “the most challenging piece of the implementation is understanding the data sources and defining the business rules,” she said. “What do you want to do with all this forecast data that’s coming in?”
- One member at the session said data from his TMS gave the treasury team a good sense of regular cash balances. The next step, he said, is following BMS’s example and leveraging data from its ERP. However, “The biggest challenge we have is wading through the data to understand what data is relevant to our forecast.”
- Ms. Lurye agreed. “That is definitely challenging. We had to sift through the data to understand what components of it we were going to use to build the logic. We also spent a lot of time understanding how our data flows in the current process and how we want to design the future state.”
- She added, “We’ve had a lot of guidance from the TIS team in terms of understanding the possibilities of the system and defining how we want to derive the cash forecast.”
- The embedded AI chatbot that TIS is developing should simplify this process, enabling treasury to define business rules and forecasting logic much more easily, Mr. Paquette said.
- “We’re working on a way to simplify that business logic, those rules, where through natural language, you could type what you want to create for a business rule and then the chatbot would give you the programming piece of it that you plug into the platform. Improvement is coming.”
Nuts and bolts. The deck Ms. Lurye presented explains the background, objectives and approach of the BMS cash transformation journey. Here are some key takeaways and insights from the presentation:
- BMS stressed the importance of taking the time to engage in “blueprinting” to understand your current process and envision the future state. Before, forecast methodologies were not consistent across the globe and relied on consolidating multiple Excel models, which risked making the forecast obsolete.
- BMS uses the TIS forecasting tool for daily forecasting of the next 12 months, drawing on data from the TMS, the ERP and a budget and planning tool. Actuals are updated daily on a one-day lag.
- Ms. Lurye said, “One of the biggest wins we have achieved so far is that we have visibility into legal entity forecasts, and the line items of that forecast are consistent across the board. We also have visibility into liquid and invested cash by legal entity.”
- BMS is working with TIS to enhance forecasting logic to incorporate daily actuals. “‘For example, if we are forecasting trade collections for one of our markets at the end-of-month, we could take into account the actuals that have come through and reduce the remaining forecast.”
- BMS also sees opportunities in a forthcoming liquidity module within the TIS tool; wants to explore monthly reporting to fully leverage forecast variance analysis in the tool; and is partnering with TIS to identify additional enhancements.
