ChapsVision AI Summit 2025
We had the opportunity to attend an AI-focused ChapsVision event in New York City. The meeting was a continuation of an annual event hosted by Sinequa by ChapsVision, one of the companies acquired to form ChapsVision, and was well attended. We were introduced to Sinequa and their AI-powered enterprise search a couple of years ago, and we’re excited to see the significant role they are playing within ChapsVision.
ChapsVision is an interesting solution provider. They are only 5 years old, but they have acquired 29 companies to form what they call “the trusted partner for your agentic AI journey.” We had the opportunity to hear Brian Kirk, GM North America, share their view on the future of the AI market. In his terms, “the future is agentic.” He shared some important ChapsVision beliefs (see image) to describe their vision.
The beliefs Brian shared align well with our view of AI, in particular, the importance of accessing company knowledge and the importance of vertical industry understanding. Our research and experience show the importance of verticalization for AI. In specific, our Making Manufacturing Analytics and AI Matter report shares that manufacturers want industry- and application-specific AI software. Solutions are more valuable when they understand the data, semantics, and the context, which is all very specific to industries. In Brian’s words, “The agentic AI war will be won in the (Industry Vertical) trenches.” We couldn’t agree more.
Orchestration Economics
The event was grounded by one of the best AI keynotes I’ve seen. We heard from Raphaelle d’Ornano about Decoding Discontinuity. It was a high value, low pretense discussion where she introduced “Orchestration Economics” and the importance of orchestrators and agents to deliver AI value. She shared three laws of economic value in the Agentic Era:
- Proximity to user intent (be the person that receives the AI request from the user)
- Access to context (provide data in the right knowledge graph)
- Coordinate the workflows (be the orchestrator)
It was a great construct to think about how agentic AI adds value and what determines solution provider worth.
ChapsVision Embraces Vertical AI Solutions
The conference went beyond high-level positioning and pillars to share details about ChapsVision’s offerings. ChapsVision offers three primary solution platforms:
- ChapsAgents – a platform for deploying and managing AI Assistants / Agents
- ArgonOS – a comprehensive data processing foundation
- Sinequa – enterprise-grade RAG (retrieval augmented generation) and enterprise search
We learned more from an education presentation by CPO Jeff Evernham, who presented on the five eras of agentic AI revolution before sharing details about their products. Some of our key takeaways were that ChapsVision is packaging agents as tools, becoming an agentic platform, and verticalizing. Further, we heard that they are committed to being multi-model and allowing customers to choose their own LLMs. ChapsVision doesn’t want to create an LLM, he explained, they want their customers to get more value from them.
Solutions for Manufacturing
As mentioned earlier, ChapsVision believes in industry specialization. They focus on key verticals, including manufacturing, energy, life sciences, private equity, and legal businesses. There was fascinating information about ArgonOS for data ops and decision intelligence, and they shared examples for manufacturing, supply chain, and maintenance. We also heard about the value of their Systran solution for language translation. Another interesting proofpoint of their industry focus was the explanation of their AI Workplace solutions. Mingee Kim , Head of Customer Success, and Lead Presales Consultant Irene Margarit showed how they are building repeatable, verticalized solutions. The current focus has been on Legal and Financial Services verticals, but the approach is applicable to all of their verticals. We look forward to learning more as they extend Workplace solutions to manufacturing.
Drill Down on Sinequa
Based on our own industry specialization in the industrial industries, what caught our attention the most were the solutions specific to manufacturing. In particular, we were excited to get a refresher and learn more about Sinequa. We and saw a Sinequa demo focused on finding part information, including specs and drawings, which are typically spread out across multiple systems. The demo explained that companies need to find part data so they don’t recreate them, and went further into the negative financial impacts of the inability to retrieve part information.
They explained that Sinequa is a governed, secure, and trustable solution grounded in the content of the organization. He shared that Sinqua connects to a variety of enterprise systems, including PLM, ERP, SharePoint, and more, but only shares what the user is allowed to see based on underlying permissions.
Miles Yaeger demonstrated a Part360 Dashboard of the part, connected and contextualized, showing a comprehensive overview of part data. To demo showcased their understanding of engineering and manufacturing data, testaments to their vertical focus, including:
- Design evolution, including prior revisions and change documents
- 3D CAD data, including the model tree, and 2D schematics
- The ability to index and visualize the BOM
All of this data was traceable back to the engineering documentation sources, providing trust while also offering users the ability to drill down to the underlying data.
Cummins Shares Engineering Search Success
One of the highlights of the event was a presentation by global power technology leader Cummins Inc.. Cummins was one of a number of larger, world-class manufacturers in attendance including Boeing, Blue Origin, Northrop Grumman, and Exxon. We heard from Scott Beard, Engineering Efficiency Project Manager for Cummins, as he shared an informative and entertaining story of how Cummins chose to use Sinequa as their solution for their 13,000 engineers to look for specialized information.
Cummins ran one of the most metrics-driven proof of concepts (POC) I’ve heard of. First, they selected 200 testers from across their business units and functions and asked them 36 questions about their ability to find and leverage product information. They gained a baseline that showed a 19 out of 100 performance score and 4.4 out of 10 on a satisfaction scale. Further, they calculated an average of 5.3 hours per week of wasted time looking for data. That is a common scenario, according to our Effective Design Data Management and Collaboration research study, that shows engineers typically spend 19% of their time on non-value-added data management activities, which equates to about one day per week.
The scoring for the POC included objectives for productivity, lifting real-world decision-making performance, how desirable the solution is to users, whether it beats incumbent solutions like PLM, and whether it scales beyond tech search. In addition, he explained, it had some advanced features test. Multiple solutions were benchmarked and Scott says that Sinequa was the clear winner. The results included a 63% productivity gain, representing a 2.9 hours per week improvement. Sinequa also performed well on other metrics, including the performance and interest scores, leading to their selection as the solution of choice. It was a fascinating story set to a theme of The Wizard of Oz, of all things!
Scott reported that Cummins achieved a 15X ROI from their investment. This confirms what Jeff Evernham said earlier in the day, that although a recent MIT study shares that 95% of companies are missing agentic AI value, the other 5% are seeing significant value.
Going forward, Cummins explained that they see more value from Sinequa than just part search. As Scott explained, they now view Sinequa as their “Knowledge Hub for Enterprise AI.” They’re looking at 7 different design patterns and planning to implement virtual SMEs (subject matter experts), such as a geartrain expert. He also pointed out the value of capturing Cummins corporate IP (intellectual property) by interviewing retiring workers to build out a Cummins knowledge base.
Our Take
ChapsVision has assembled a significant collection of AI capabilities delivering vertical solution value. We were impressed with our earlier conversations with Sinequa, and now we are more excited to hear about how ChapsVision will leverage and extend Sinequa’s capabilities for the manufacturing industries. We’re excited to see how the combined offerings help industrial companies get business value from AI and their digital threads, and look forward to watching the progress with workplaces to deliver more ready-to-use, verticalized solutions.
Thank You
Thank you, Laurent Fanichet, for inviting us to the event and to Stacey Greene and the team for their organization that made it such a valuable time. Further thanks to Brian Kirk, Jeff Evernham, and Xavier Pornain for sharing your vision and progress in AI for the manufacturing industries. We look forward to staying in touch and learning more about your plans.
