Coveo enhances AI for Agentforce with context passthrough tools
Coveo has introduced two new capabilities-custom context passthrough and broader compatibility-designed to enhance the performance of AI-powered support experiences on Agentforce.
The company announced the enhancements as part of its ongoing efforts to address the enterprise challenge of siloed data and inconsistent knowledge, which often hinder the effectiveness of AI agents in delivering precise, trustworthy results.
Feature update
With the addition of custom context passthrough, Coveo's Relevance-Augmented Passage Retrieval API (PR-API) can now tailor information retrieval using highly specific user data, such as agent role, account tier, or product line. This approach aims to provide more accurate and contextually relevant responses, with the goal of improving self-service success, shortening handling times, and reducing mean time to resolution.
Joyce Leung, Vice President of Support Operations at Illumio, described the impact of combining Coveo and Agentforce on internal support testing:
"When we combined Agentforce with Coveo, the results were on another level. Our internal testers went from saying, 'this is okay' to literally saying 'wow.' The multiplier effect of bringing the two solutions together was incredible, consistently returning responses that truly amazed our team."
The integration extends to allow enterprises using Agentforce and Coveo to move beyond simple question-and-answer pairs. The platform supports a range of GenAI-powered functions, including autonomous case resolution, auto-drafting emails, and generating knowledge articles. All actions are based on enterprise-approved content.
Industry expectations
Organisations in the fields of service, commerce, and marketing continue to seek efficiencies in delivering faster and more trusted customer experiences at scale.
Jujhar Singh, President and General Manager, Applications and Industries at Salesforce, commented on the partnership:
"Service, Commerce and Marketing leaders everywhere are under pressure to deliver faster, more accurate, and more trusted experiences at scale. Combined with Agentforce, Coveo supports the next generation of AI agents that meet that challenge by combining deep enterprise knowledge with contextual intelligence, empowering teams to serve customers with precision, speed, and confidence."
Coveo says that its improvements aim to help organisations reduce cost-to-serve, accelerate time-to-resolution, and improve both self-service and assisted service operations. The unified index and context-first approach are intended to address ongoing concerns in generative AI deployment such as accuracy, hallucination risks, security, and content trust.
Enterprise focus
The announcement builds on Coveo's strategy to support enterprises as they move from experimental generative AI deployments to wider implementation of AI-driven agents within established business processes.
Richard Tessier, Co-Founder and Senior Vice President of Product at Coveo, said:
"With these capabilities, we're giving enterprises key pieces to build successful agents: control and context. Coveo for Agentforce goes beyond better answers - it enables smarter actions, grounded in enterprise truth, shaped by the user's reality. That's how you move from experimentation to agentic transformation."
The platform processes queries sent from Agentforce, returning results from connected content sources, all within the framework of enterprise content approval. This is aimed at ensuring that the outputs of agentic or generative AI actions remain secured and compliant with organisational standards.
As enterprises continue to deploy AI solutions, many are prioritising improvements in personalised support and customer interaction. Contextual AI capabilities such as those announced by Coveo may assist businesses in bridging the gap between fragmented internal knowledge and the demands of responsive customer engagement.
The new features are included in the Coveo AI-Relevance Platform and are available for enterprises seeking to extend their use of AI within support and operational environments.
