After selling his latest startup to IBM, this founder now wants to bring ‘agent memory’ to enterprise data

Not everyone agrees on what a “I have an agent“Actually it is, but they They are all still in fashion. At a broad level, these so-called “agents” promise to go several steps beyond a simple chatbot, making decisions and taking actions on behalf of people. Some could help you do your shopping online; others could make factories more efficient — with different degrees of autonomy.

It is in this context that the nascent AI startup, Through AIis coming out of stealth to develop a “dynamic memory system” for complex enterprise workflows. And it’s headed by a founder who relatively recently sold his previous startup to IBM.

Across AI targets people like chief revenue officers and sales teams with a platform that connects to all of their internal and external business data sources. It then creates a shared “agent memory” that can be used to identify and qualify new sales opportunities, detect risks, and suggest questions that sales teams should ask their customers.

“Sales teams often struggle to get and use the right information when they need it, whether it’s information about products, customers, competitors, or optimal processes,” said Across AI co-founder and CEO, Steven Michael (pictured above, center), told TechCrunch. “Critical knowledge is often locked among a few experts or buried in large amounts of unstructured data, leading to inefficiencies, delayed decisions and costly errors. “Existing AI solutions often fail to address this problem because they lack deep integration and contextual understanding, and treat all data equally without the ability to prioritize or adapt to new information.”

Mih was previously co-founder and CEO of Punishmenta company backed by Google Ventures that created commercial services on Prestothe open source SQL query engine that It split from Facebook in 2013.. mih sold Ahana to IBM last year for an undisclosed amount, and after a 14-month stint at the tech giant, Mih jumped ship in July to begin working on his latest startup.

Joined Dr. Niloufar Salehi (pictured above, left) and Dr. Afshin Nikzad (pictured above, right), renowned professors from UC Berkeley and Stanford University, respectively, who have conducted investigation on ways to improve the effectiveness of AI systems in “high-risk” environments.

Across AI is still in an embryonic state: it is refining its product with design partners privately. As it works toward a commercial launch in 2025, the company has raised $5.75 million in a seed funding round co-led by Bobby YazdaniCota Capital, and Global Villagea venture capital firm that counts Bill Gates, Mark Zuckerberg, Jeff Bezos and Reid Hoffman among its sponsors.

Personal opinion:

“After selling his last company to IBM, this founder seems to be looking to provide smart solutions in the field of artificial intelligence. If he succeeds in this goal, it could contribute to improving data management and providing deeper insights by enabling systems to remember past contexts and interactions. This technology could change the way companies deal with their data and increase the efficiency of intelligent systems in decision-making.”

Building memories

Through AI there will be a web application and a chatbot that will connect to various parts of the business stack (CRM systems, communication and collaboration tools, calendars and everything else) to build its memory and develop contextual understanding. It will then be available to help wherever a user is working.

“By showing up where users already are, for example in Slack or the (Microsoft) Team app, Across AI apps don’t disrupt user flow and instead provide just-in-time assistance in the context of workflows.” user’s existing data,” Mih said.

This memory, the company says, “continuously adapts” and only retains what it considers relevant information while discarding obsolete data. This raises questions about how you could determine what is relevant, as this depends largely on the context and the requirements of the people who will use it.

Mih says he does this by developing a “deep understanding of the context of the workflow.”

“The system actively tracks, timestamps and monitors information updates, recognizing when data becomes outdated or conflicts with new information,” he said. “Unlike traditional AI systems that treat all data equally, our agent memory system prioritizes information based on contextual importance. Whenever possible, applications keep inferences up to date on their own. When ambiguity exists, determinations are escalated to a relevant person, such as a sales manager or product manager.”

Of course, companies have been slow to adopt Generative AI, as data privacy and security remain key concerns. The last thing a company wants to do is funnel all of its confidential and proprietary data to a third party, who then does God knows what with it. As such, Mih says data security is a “fundamental aspect” of the startup’s agent memory platform.

“Our memory system operates within the enterprise’s secure environment, maintains access control to sensitive information, and does not expose data to external models for training,” Mih said. “We plan to offer both SaaS and on-premises cloud deployment options to meet enterprise security and compliance requirements.”

There are subtle synergies between Mih’s previous startup and his latest venture. Ahana was about allowing users to query large amounts of data through Presto, and Ahana took care of all the complexities related to setting up and maintaining the infrastructure. Across AI addresses the same problem, but through a different lens.

“I think a core differentiator for AI application companies will be their ability to help users analyze large amounts of data quickly – that’s exactly what we specialize in at Ahana,” Mih said. “This experience deepened my understanding of the challenges businesses face in making sense of complex data ecosystems that are often siled and difficult to navigate.”