Salesforce has signed a deal to acquire Spindle AI, an analytics startup focused on agentic AI. It’s the latest in a series of moves towards Salesforce fulfilling its vision for the “agentic enterprise” through its Agentforce platform.
Spindle AI’s technology “combines advanced AI agents and machine learning with powerful data modeling to help businesses make faster and more effective data-driven decisions,” according to Salesforce’s announcement. It complements analytics platforms like Tableau by using AI to model agentic scenarios autonomously and forecast business outcomes.
This is designed to tackle a familiar problem for enterprises producing massive amounts of data. By simplifying analysis and interpretation, they can unlock that data’s value and generate insights to guide critical decisions.
Spindle AI was co-founded by Ryan Atallah, who previously sold his startup ClearGraph to Tableau, and Carson Kahn, who built the AI company Volley ML during JPMorgan’s In-Residence program.
The team brings experience in multi-agent systems and high-performance analytics, which are areas Salesforce clearly sees as key to helping its AI agents move beyond assisting humans to autonomously analyzing data and optimizing decisions.
Multi-agent orchestration is the next phase of AI development, where multiple specialized agents work together by communicating, reasoning and coordinating tasks. This can enable enterprises to advance their customer experience platforms from simple chatbots toward adaptive, end-to-end autonomous operations.
Adam Evans, EVP & GM, Salesforce AI Platform, framed the deal as part of the company’s larger vision:
“Bringing Spindle AI—which includes veteran leaders and incredibly talented engineers with deep expertise in advanced agentic analytics and machine learning—onboard will continue our investment in Agent Observability and Self-Improvement to help us deliver custom agentic analytics, ROI forecasting, and continuous optimization for every Agentforce user.”
Spindle’s team will join the Agentforce division to strengthen capabilities like agent observability, which tracks how agents reason and perform, and self-improvement, helping them to learn from outcomes.
How Salesforce Is Building Its Agentic AI Ecosystem
The Spindle acquisition fits neatly into Salesforce’s move towards making AI agents core to its platform.
Earlier this year, the company announced its acquisition of Informatica, a data-management company that provides tools to support the clean, governed data pipelines needed to power AI agents.
Salesforce also acquired Convergence.ai, a startup focused on developing adaptive AI agents that can navigate complex workflows and dynamic interfaces, adapting in real time to challenges like pop-ups, errors and UI updates.
These buys highlight the company’s ambition to own every layer of the agentic stack, from data quality and observability to reasoning and optimization.
Spindle fits into Spindle fits into this strategy by adding advanced agentic analytics and scenario modeling capabilities. Teams can test out pricing changes, campaign shifts, or service workflows before they happen. That kind of forecasting could make AI agents more proactive and strategic, not just reactive.
“We built Spindle AI to intelligently close the gap between what questions enterprises want to ask of their data and what their data systems can understand,” Ryan Atallah, co-founder and CEO, Spindle AI, explained.
The aim is to integrate Spindle’s capabilities into Agentforce so that agents become more reliable and better at justifying their recommendations. Carson Kahn, co-founder and Chief AI & Product Officer at Spindle AI, added:
“We look forward to accelerating Agentforce with sophisticated agentic analytics and forecasting that make enterprise LLMs more reliable and valuable.”
Salesforce’s pitch for Agentforce is that enterprises need AI agents that don’t just automate tasks but understand and improve how they work. As Jayesh Govindarajan, EVP Salesforce AI, Agentforce, put it in the announcement:
“Ryan, Carson, and Spindle AI have proven expertise in the complex AI observability and multi-agent analytics functions that are critical for measuring and forecasting AI-driven value.”
That should translate to systems capable of self-analysis—monitoring performance, adjusting strategies, and providing transparent reasoning behind every recommendation.
Still, questions remain about how quickly customers will adopt these more autonomous tools, and how companies will manage governance, explainability, and trust at scale.
The Spindle AI acquisition is expected to close in Salesforce’s fourth fiscal quarter of 2026, subject to conditions. Once complete, it will mark another stage in Salesforce’s effort to turn its customer platform into an ecosystem of intelligent, continuously improving AI agents.
