Seismic

Solution

Seismic worked with IBM's Ecosystem Engineering Build Lab team to create a minimum viable product that improved how meeting transcripts are processed, summarized, and turned into recommendations inside the Seismic platform. IBM watsonx.ai supplied the generative AI capabilities used to chunk and analyze transcript files, enabling faster prototyping and more seamless embedding of AI into the existing product. The build team also helped Seismic write Terraform code so the enhanced summarization service could be deployed more consistently, and ran knowledge-transfer sessions so Seismic engineers could sustain future projects themselves. The resulting implementation linked transcript processing to the collaboration tools customers already use, including Zoom, Microsoft Teams, and Webex, which made the feature operationally useful rather than a stand-alone demo. Seismic therefore upgraded a promising GenAI feature into a more scalable, production-oriented content-processing service.

Data flow

Meeting transcript files are chunked and processed by watsonx.ai, the generated summaries and recommendations are embedded back into Seismic's platform, and the service integrates with collaboration tools such as Zoom, Teams, and Webex.