Vectara has launched the Tool Validator, its second Guardian Agent designed to enforce real-time governance in enterprise agentic systems. This innovation addresses a critical gap in AI workflows by reviewing and correcting proposed tool calls before execution, preventing costly errors, security risks, and compliance issues that plague 95% of failed AI deployments.
As enterprises rush to deploy AI agents, early workflow errors—such as incorrect tool selection or irrelevant calls—cascade into hallucinations, wasted resources, and real-world damage. Vectara's Tool Validator intervenes at the planning stage: it analyzes the agent's proposed action plan, identifies flaws, and prompts recalibration. The result? Agents execute safer, more efficient workflows with built-in traceability.
Vectara co-founder and CEO Amr Awadallah said, "While 2025 began with tremendous enthusiasm for the promise of AI agents, many organizations have since come to realize what Vectara has known all along: that without an overarching 'operating system' to ensure accuracy, reliability and security across all of an organization's agents, it is almost impossible for an agentic system to deliver on its promises or deliver the ROI that enterprises demand."
Vectara Chief Product Officer Eva Nahari said, "According to our new Guardian Agents Benchmark, hallucinations or erroneous tool calls early in an agentic workflow pose a much higher risk to enterprises than a simple wrong answer from a chatbot. Tool calls have real-world impacts: they might initiate a payment, update a customer record, or take action in a production system - not just generate a response. When we tested common open-source agent frameworks, we found that tool-call accuracy was highly variable, ranging from roughly 5% to 59% depending on the task and configuration of the system. In other words, even widely adopted agent building approaches frequently end up choosing the wrong tool or calling tools unnecessarily. This creates real operational and security risks if those calls aren't monitored and governed."
Nahari concluded, "Our new Guardian Agent: Vectara Tool Validator gives enterprises the increased confidence they need to move high-ROI agentic workflows into production by automatically adjusting agents' mistakes before they produce high-visibility, detrimental outcomes, while also ensuring visibility and traceability. This enhanced level of reliability is critical in the emerging agentic era, as AI agents will not only be assisting enterprises with crucial tasks, but will also be making critical decisions on their behalf."
Vectara is the Agent Operating System for trusted enterprise AI: a unified Agentic RAG platform with built-in multi-modal retrieval, orchestration, and always-on governance. Vectara's platform can be deployed on-premises (air-gapped), in a customer's VPC or as a SaaS solution. Vectara Enterprise-grade agents deliver grounded answers and safe-guarded actions with source citations, step-level audit trails, fine-grained access controls, as well as real-time policy and factual-consistency enforcement, so teams ship faster with lower risk, at scale.
TechIntelPro empowers B2B technology decision-makers with the latest industry trends, deep insights, leadership perspectives and actionable intelligence to drive both organizational and professional milestones—and thrive on the tech pulse.