Last year, we launched Job Architect — our first AI agent, built to help HR teams create and refine job frameworks through natural conversation. It proved that agentic AI could deliver real, practical value in HR. Not as a chatbot novelty, but as a tool that genuinely understands the structure of work.
Today, we’re launching Prism, the AI Matching Agent that provides a conversational AI experience built directly into the 365Talents platform. It doesn’t just suggest jobs. It starts a real conversation with each employee, understands their aspirations, and delivers personalised, explainable recommendations with built-in skill gap analysis.
Like everything we build at 365Talents, Prism was shaped by the conversations we have every day with our clients. Their challenges, their feedback, their ambitions for their people are what drive our product forward. We don't build for them, we build with them.
This is the next step in a journey of continuous AI innovation at 365Talents. Live and available now for every 365Talents client.
From suggestions to conversations
Profile-based matching delivers real value. It surfaces relevant roles quickly, at scale, with no effort from the employee. But there's an inherent ceiling to the model: it can only work with what it already knows about you. It can't ask what you actually want.
An employee might be ready for a management role. Another might want to do the same job in a different city. Someone else might be open to a change but not sure what direction to take. These aspirations are nuanced, personal, and constantly evolving — and no static list can capture them without being told.
The question was never "how do we fix suggestions." It was "how do we give every employee a way to tell us what they're looking for, and respond intelligently?" That's what the Prism does.
A new approach: matching through conversation
The Prism Agent reimagines what the matching experience can be. Instead of presenting a static list of suggestions, it opens a dialogue.
An employee might say: “I want to grow into a management role,” or “I’d like to do the same job but in another city,” or “I want to change direction but I’m not sure what’s out there.” The agent takes it from there — asking clarifying questions, analysing the employee’s profile against available opportunities, and presenting recommendations that are tailored to what they actually expressed.
But recommendations alone aren’t enough. What makes the Agent different is what comes with each suggestion:
- Explainability. Every recommendation includes a clear rationale. Employees see exactly why a role was suggested based on their skills, experience, and stated aspirations. No more black-box matching.
- Skill gap analysis. When a role is a stretch, the agent doesn’t simply flag it as a poor fit. It identifies the specific skills that are missing and, where possible, recommends training to bridge the gap. Employees see a path forward, not a dead end.
- Contextual intelligence. The agent respects your organisation’s matching rules and configuration. It operates within the boundaries you’ve set, ensuring that suggestions are not only relevant to the employee but aligned with your mobility policies
How the AI Matching Agent works
The experience is designed to be intuitive. No training required. No complex onboarding. Here’s how it flows:
- Step 1: The employee starts a conversation. From the homepage or the opportunities tab, employees can access the agent and describe what they’re looking for in natural language. Starter prompts are available for those who prefer guided entry points.
- Step 2: The agent analyses and recommends. Drawing on the employee’s profile, skills, and the conversation context, the agent identifies relevant open positions and job roles. Each recommendation comes with an explanation of why it was selected.
- Step 3: Gaps are identified, actions are taken. For roles where there’s a skill gap, the agent provides an honest assessment and suggests training. Employees can view full details, save opportunities, or apply — all within the conversation flow.
Conversations are private. They are not accessible to managers or HR, and the AI does not train on these exchanges. 365Talents is ISO 42001 certified, and all data handling follows our existing security framework.
What the AI Matching Agent is, and what it isn’t
Transparency matters, so let’s be clear about what this agent does and doesn’t do.
It is a conversational tool that helps employees explore their immediate next move. It surfaces relevant opportunities, explains its reasoning, and identifies skills to develop.
It is not a long-term career coach, a replacement for HR, or a magic solution that guarantees perfect matches. It depends on the quality of data in your platform — the richer your job postings and employee profiles, the better the agent performs.
It complements the existing suggestions tab. It doesn’t replace it. Employees can use both. Think of the agent as a guided, conversational path alongside the traditional browse-and-filter experience.
Part of a bigger story: AI that keeps evolving
The Prism Agent doesn’t exist in isolation. It’s the latest milestone in a deliberate, continuous innovation roadmap that has been accelerating over the past two years.
In 2024, we shipped LLM Extraction V2 — a major upgrade to our skills extraction engine. By leveraging large language models, we dramatically improved the accuracy and depth of skills data extracted from profiles, documents, and collaboration tools. Better data means better results across every feature in the platform.
In early 2025, we launched Location Preferences — giving employees the ability to express location preferences and ensuring that matching respects them.
Later in 2025, we introduced Job Architect — our first AI agent. Job Architect helps HR teams create, refine, and visualise job frameworks through natural conversation. It proved that agentic AI could deliver real value in HR, not as a chatbot novelty but as a tool that understands the structure of work with genuine HR expertise.
And now, in 2026, the Prism Agent extends the agentic model to the employee experience. It’s our second AI agent, and it won’t be our last.
This pace of innovation is deliberate. At 365Talents, AI is not a feature we bolt on to check a box. It’s the engine at the core of the platform, and it evolves with every release. When you choose 365Talents, your platform doesn’t stand still. It gets smarter, quarter after quarter.
What this means for HR and mobility teams
For HR leaders and mobility managers, the AI Matching Agent addresses a practical challenge: how do you scale personalised career conversations across an entire organisation?
The answer, until now, has been that you can’t. HRBPs and career advisors can only engage in a limited number of in-depth conversations. The rest of the workforce gets a suggestion list and a link to apply. The Prism Agent fills that gap. It handles the exploratory phase — the “I’m curious, what’s out there?” conversations — at scale, 24/7, in any language the platform supports.
This doesn’t replace the HRBP. It amplifies them. The agent takes care of the straightforward cases so that your human experts can focus on complex mobility scenarios, strategic talent moves, and the conversations that truly require human judgment and empathy.
For organisations that have struggled with employee satisfaction around matching quality, the agent also represents a tangible response. When your next employee NPS survey asks about internal mobility, you’ll have something concrete to point to: a conversational experience that listens, explains, and guides.
How to get started
The Prism Agent is available now for all 365Talents clients. You can activate it from your Configuration Centre in a few clicks. No heavy setup, no training programme required.
We’ve prepared a set of resources to support your rollout.
Your Customer Success Manager can also schedule a personalised walkthrough at any time.
Not yet a 365Talents client? Request a demo to see the AI Matching Agent in action and learn how 365Talents is building the future of internal mobility with agentic AI.
