One of IBM Process Mining's primary goals has always been to make digital twin process analysis easy and swift. Creating dashboards populating necessary data has never been an issue, thanks to Analytics and its simple widgets. With the next version we're taking this to the next level.
Introducing: Data analyst.
Envision an on-demand SQL expert, available at any time, simplifying queries about your process statistics without the need for technical knowledge. The groundbreaking text-to-SQL feature within IBM Process Mining’s latest chat interface now allows for the fast inquiry and reception of critical analytics, such as "identify the activity with the shortest duration and highest cost over the previous month", instantly.
Beyond just data retrieval, this innovative "Data Analyst" assistant also facilitates data visualization via its text-to-chart capability. Not only does it extract relevant data but also organizes it, enhancing accessibility and enabling easy export into downloadable, shareable chart formats.
Moreover, for advanced users, the Data Analyst goes a step further by displaying the actual SQL instructions, achieving two significant objectives:
1. Building trust and transparency, users gain insight into the precise methodology behind the results.
2. Empowering Analytics integration, the generated SQL queries can serve as foundational elements in Analytics for creating, modifying, and refining widgets, adapting analytical dashboards to specific needs.
The strength of this new assistant lies not in being a simple "one-shot" text-to-SQL translator, but rather in its comprehensive grasp of the domain and data structure within IBM Process Mining. This capability is realized through a specialized fine-tuning of the Granite AI model, facilitated by a collaborative effort with the FlowPilot team. A key feature of this system is its capacity to incorporate foundational knowledge from diverse sources, resulting in a high level of dependability and precision.
We are confident that the "Data Analyst" feature will enable users to access the extensive data and insights extracted by IBM Process Mining more quickly, facilitating both in-depth and expeditious process analysis. We are excited to make this functionality available to everyone in the next version!