The funding will probably be used to broaden engineering and speed up enterprise deployments
ChatSee.ai, which gives the failure intelligence layer for autonomous AI programs, introduced a $6.5 million funding spherical led by True Ventures, with participation from First Rays Enterprise Companions, Seven Hills Ventures, and business veterans.
Autonomous AI brokers are quickly transferring from experiments to manufacturing in enterprises. Customized brokers constructed on high of fashions from OpenAI, Gemini, and Anthropic—together with embedded brokers in platforms equivalent to Microsoft 365 Copilot, Salesforce Agentforce, Snowflake and Databricks Agent platforms—are more and more powering buyer interactions, operational workflows, analytics, and enterprise decisioning programs. On the similar time, builders are constructing more and more complicated, autonomous multi-agent programs utilizing frameworks equivalent to LangChain, Microsoft AutoGen, and rising open tasks like OpenClaw.
However as these programs transfer into manufacturing, a brand new confidence hole is rising: brokers that seem succesful throughout testing typically exhibit recurring behavioral failures as soon as deployed into real-world environments. Not like conventional software program failures, many AI failures depend upon context, intent, coverage interpretation, and enterprise outcomes, making them troublesome to detect by static guidelines or standard monitoring alone.
Observability instruments assist people examine particular person agent interactions, however they don’t protect the failure intelligence wanted for programs to be taught from recurring errors.
Enterprises want a strategy to seize the context surrounding behavioral failures, perceive how they had been remediated, and decide whether or not comparable points proceed to recur. With out this organizational reminiscence, brokers can not successfully be taught from prior failures, inflicting the identical errors to recur throughout interactions, workflows, and enterprise processes—from missed escalation triggers and unintended disclosures to incorrect coverage selections, software misuse, workflow drift, and breakdowns throughout long-running operational processes.
“Most of the most important AI dangers emerge at runtime as brokers function autonomously,” stated Dr. Eduard Amoroso, CEO of TAG-infosphere (and former CISO of AT&T). “As a result of these programs are probabilistic and adaptive, static testing alone is inadequate. That is driving the necessity for steady runtime assurance throughout enterprise workflows, with platforms like ChatSee serving to organizations observe and enhance AI habits over time.”
Additionally Learn: AiThority Interview with Matej Bukovinski, Chief Know-how Officer at Nutrient
Trade analysts at Gartner® have recognized the necessity for a brand new management aircraft, known as Guardian Brokers, centered on observing and defending these programs. ChatSee was just lately included within the Gartner Market Information for Guardian Brokers within the enterprise alignment and final result optimization class, which we imagine highlighted the rising want for applied sciences that monitor and align the habits of manufacturing AI brokers with enterprise outcomes.
ChatSee is co-founded by serial entrepreneur Sekhar Sarukkai, who co-founded Skyhigh Networks (acquired by McAfee), Securent (acquired by Cisco), and Confluent Software program (acquired by Oracle). He’s joined by co-founder Sanjay Agrawal, PhD (Stanford), whose analysis and engineering work has centered on large-scale distributed programs and enterprise AI infrastructure.
“Once we began analyzing agent failures, we realized the issues appear chaotic however truly fall into repeatable patterns,” stated Sarukkai. “That’s the place observability falls brief—it exhibits what occurred, however not whether or not the habits was truly appropriate. We’re discovering that these failures fall into repeatable patterns that may be labeled, remediated, and constantly fed again into each human and AI workflows so programs be taught and enhance over time. This shifts AI operations from people merely supervising brokers to people and brokers collaboratively enhancing outcomes, turning reactive oversight into steady, ruled AI operations at scale.”
Taking a first-principles strategy, ChatSee introduces a failure intelligence layer for enterprise AI programs. Whereas observability platforms assist groups monitor what brokers do, ChatSee focuses on understanding behavioral failures, preserving the context surrounding them, capturing remediation data, and monitoring recurrence over time.
The result’s a shared failure reminiscence—a constantly rising organizational document of what failed, why it failed, the way it was mounted, and whether or not it occurred once more. This permits enterprises to maneuver past investigating failures one interplay at a time and constantly enhance how AI programs behave in manufacturing.
“AI brokers are rapidly turning into operational infrastructure inside enterprises,” stated Puneet Agarwal, Companion at True Ventures. “However corporations nonetheless lack instruments to know when these brokers behave incorrectly in manufacturing and tips on how to appropriate these failures at scale. ChatSee is addressing this vital hole within the rising AI stack.”
