Agentic AI Moves Into Everyday Work as European Startups Raise €7.9bn

Startups are driving the innovation across a wide range of sectors. They include Swedish legaltech startup Legora, healthtech Tandem Health and UK-based fintech Cleo. But integrating agentic AI models into everyday work is not simply a matter of choosing a model. For an AI agent to execute tasks, it must be able to interpret often messy and unstructured data from internal systems as well as the web.

Emma Burrows, cofounder of Rezonant, which develops software for organising a company’s data before it begins automating work, says building a reliable agentic workflow requires turning that data into a structured “perception layer” that agents can reason with. This layer involves training AI models to “organise context from a lot of different inputs,” she says, ranging from communication platforms such as Slack and email to documents and files.

Rezonant supports companies by building a tailored “context graph”, which maps scattered information and data into connected business categories. Burrows says the context is not organised by every document put together, but categorised by features or customer insights. By structuring data around categories within the business, such as sales, operations or finance, agents can pinpoint exactly what data they need to make a decision.

The next phase for enterprises is owning how their company-specific intelligence improves. Many organisations mistakenly believe they need to modernise their entire tech stack before they can think about using AI. Lindsay Keim, VP of customer success at AI agent startup N8n, warns that waiting to organise data stacks often prevents companies from starting at all, leaving them stuck in a “data cleaning purgatory.”

N8n is an AI platform connecting LLMs, data sources and business tools. Keim says the flexibility of platforms like N8n allows users building agentic workflows to “pull data points out of a legacy system, run them through a separate AI model for analysis and move them into a newer system, without actually modernising the whole stack first.”

When workflows require external data, challenges often come from the unpredictability of the internet. Rotem Weiss, founder and CEO of agentic search company Tavily, which provides an API connecting AI agents to web information, says agents need an “abstraction” layer to “turn the messy, unpredictable web into consistent context.” Without that layer, he says, an AI agent struggles to find and process reliable and relevant information from the web.

Giving AI complete autonomy to execute tasks across business systems that may be sensitive or confidential can introduce operational and compliance risks. Organisations are increasingly prioritising trust boundaries, which separate a company’s internal data from the public internet, as well as governance and human-in-the-loop safeguards. Burrows says the permissions context is “the most difficult and important part” of what Rezonant does, referring to the permissions users can give an agent when it makes decisions on their behalf. Each workflow available for an agent has specific permission controls that can be configured on top of it, she says.

Agents that initially rely heavily on human feedback can “over time, learn routing patterns,” Burrows adds. She gives the example of pinging a colleague, Sam, whenever there is a decision about design. Over time, the agent may notice that and ask whether to ping Sam in future when the same situation arises. The agent can start to take on more work the way a user would, she says, but still keeps humans in the loop for high-stakes decisions.

Weiss says that when agents retrieve information from the web, external inputs are not always shared each time, so teams need to ensure caution is built in. “When an agent needs information from the internet it should send only the minimum context necessary [from an organisation’s data] to retrieve that information,” he says. Governance also requires strict auditability, according to the Sifted report.