Kantata delivers One Smart Agent for Every Project and Engagement

Kantata has unveiled the Enterprise Agent. One of the latest release features of the Kantata Expertise Engine that it announced last year. The company also announced embedded generative business intelligence, self-executing workflows, and a proprietary PSA knowledge graph.

It intends to turn its leading Professional Services Automation solutions into a tool that delivers autonomous resource planning and project management. It aims to lighten the manual workload for professional services firms.

The Expertise Engine is not a specific agent for a specific task, an approach that most other vendors are taking. Instead, it is an agent that understands the data and processes within the PSA platform and can automate these using natural language prompts.

Michael Speranza, Chief Executive Officer, Kantata. “Services organizations don’t need more AI features layered on top of disconnected systems. They need a system that understands how their business works and can act on that understanding.

“The Expertise Agent and new agentic capabilities we’re delivering with the latest release of our Expertise Engine are that system and will help teams move from execution to more consistent, intelligence-driven operations where every project benefits from what the business has already learned.”

In a soon to be released interviews, Speranza highlighted that other firms have focused on delivering an Agent Catalogue of 10, or even 50 agents. The Kantata approach is much broader, building an agent that will deal with a cross-section of industry processes.

The Expertise Engine delivers the complete context both from the PSA solution, and within, using the data relating to the clients of PSOs. This means that the Expertise Agent is fully knowledgeable about how a business works and the customers it works with.

The Expertise Agent

The Expertise Agent in Kantata is a single, unified AI assistant that adapts to where you are in the platform and what you’re trying to do. Rather than having lots of separate “project” or “resource” agents.

When you ask it to do something (for example, “create a staffing plan with three scenarios”), it:

• Understands context (e.g., the specific client/project you have open).
• Assembles the right skills behind the scenes (such as the resource optimizer that scores and recommends resources).
• Generates options and trade‑offs, like multiple staffing scenarios balancing skills, availability, and cost.
• Takes direct action in the system. Such as updating placeholder resources, assigning team members, or posting briefs to the internal project feed.
• Builds sales‑to‑services handover briefs from call logs, emails, and meeting summaries.
• Drafts change orders and customer emails from transcripts.

In addition, a new Agent Studio enables users to deliver specific instructions to the Expertise Agent in natural language. It allows users to modify how the agent works, ensuring, for example, that if the user creates a works order, that work order is structured in a specific way.

There is also a GUI where users can edit workflows. In addition, when combined with workflow automation, the agent can run proactively off triggers (status changes, schedule, API calls), Therefore, surfacing recommendations before users even log in.

Rather than relying on narrow resourcing agents, firms can state objectives in natural language and use the agent to:

• Catch red projects early – Monitor signals in real time to reduce reporting bias and prevent margin erosion.
• Staff teams faster – Match resources by skills and capacity, making resourcing more strategic.
• Smooth handovers – Preserve institutional knowledge and generate plans and briefings so new team members ramp quickly.

Architecture

Underpinning the knowledge agent is technology such as a knowledge Graph and an MCP connector. The Expertise Engine’s knowledge graph connects data from projects, people, systems, documents, communications, and meetings. Built for professional services, it maps relationships between skills, outcomes, and delivery patterns to improve recommendations and actions over time.

In addition, as Kantata ingests other data from third-party systems, the knowledge graph catalogues the information. And makes it available to the agents in the right context.

Within an open MCP-based orchestration layer and open APIs, customers can also connect to different LLMs. Kantata has built 161 tools/connectors enabling LLMs to connect to Kantata OX and later Kantata SX.

Agentic Business Intelligence

Kantata’s new agentic business intelligence turns raw operational data into proactive, AI‑driven insights and outputs. Instead of static dashboards you have to pull manually. Data visualisations, once created, are interactive. They can be drilled down into, and the user can ask searching questions in natural language to surface deeper insights.

Using the MCP connector, BI agents can access a wide set of read/write tools (everything except delete) across projects, resources, time, and financials.

Users can ask natural language questions like “show me all time entries with rejections and reasons” or “summarise the most recent project created” and get structured, interactive tables back with filters, counts, and statuses (e.g., rejected vs pending entries). It can also perform root cause analysis of any anomalies detected on request.

Because the BI layer is agentic, it doesn’t just visualise – it can also take action wherever the API allows: drafting follow‑up messages, preparing approvals, or performing bulk updates in response to what it finds. It leverages the LLM’s document‑generation capabilities to create rich artefacts. Such as executive status report slide decks assembled automatically from live project data, including scope, progress, risks, and financials.

All of this runs in the tools people already use (including Claude, Chat GPT and Gemini), reducing friction and dramatically cutting the time needed to go from “question” to “insight” to “action” – while keeping enterprise controls around what the agent can and cannot do.

The MCP server will also enable the Expertise Agent to connect with other agents to coordinate processes across systems such as finance, CRM and ERP platforms.

Is this a new era for professional services automation?

With this announcement, Kantata offers a new argument for success. It believes that Professional services firms need more than analytics tools or single-purpose agents. They need an operating model that embeds intelligence into day-to-day work.

Kantata’s Expertise Engine uses AI to learn from successful delivery patterns, identify risks earlier. And turn that knowledge into coordinated action across the business. Problems that once took months to detect and address can now be surfaced and acted on in minutes.

Agentic BI flags at-risk projects and recommends mitigation strategies. While the Expertise Agent uses project data, meetings, documents, and proven behaviours to create tailored action plans. Its conversational workflow capabilities also help convert today’s lessons into reusable institutional knowledge for future projects.

Vikas Nehru, Chief Technology Officer, Kantata, commented, “When it comes to AI, what matters is not just getting answers faster, but improving how the business operates. By combining a knowledge graph, agentic intelligence, and workflow orchestration, we’re helping services organizations turn their experience into a system that drives better decisions and more consistent outcomes.”

Looking forward, Kantata will introduce accelerators which deliver, in Speranza’s words, a “cockpit that helps them make better, smarter, more informed decisions. For example, a service accelerator would be a cockpit that a project lead or project team sits down in front of every single day, and it brings context, reasoning, and next best action to them in one single place.”

It is also developing an internal consultant cockpit to help individual consultants use the PSA in a better way. Which makes them more productive and enables them to make better-informed decisions. Perhaps using insights from other projects they were not involved in. This approach is what Kantata aims to turn organisations into learning organisations. Where corporate knowledge is never lost and is available immediately across all employees.

Enterprise Times: What does this mean

For professional services firms, Kantata is giving a glimpse of what the future will look like. While Kantata argues that its approach is unique, other vendors are similarly approaching the result. Klient, Rocketlane and Certinia recognise the importance of domain expertise. While they are not at the same point of having a general expertise agent, they are moving towards the approach at different speeds.

For Kantata partners, these new agentic capabilities open up a richer, higher‑value services opportunity. The Expertise Agent, Agent Studio, workflow automation, MCP connector and agentic BI together give partners a powerful, extensible platform to solve long‑standing customer pain points around resourcing, sales‑to‑services handoff, change control, time approvals and reporting.

Instead of one‑off customisations, partners can design reusable, tailored agents, workflows and BI experiences that embed their domain expertise directly into clients’ day‑to‑day operations. This not only deepens partner stickiness and recurring services revenue but also positions them as strategic AI advisors. Rather than simple implementation providers.

What is missing from the Kantata announcement is detailed pricing information and early customer feedback. Hopefully, Kantata will soon be able to share anecdotes and perhaps a case study with real metrics about how a professional services firm has evolved.

The post Kantata delivers One Smart Agent for Every Project and Engagement appeared first on Enterprise Times.