Recruitment firms are not short of candidate data. The harder problem is that much of what recruiters actually know about a person gets buried inside calls, emails, notes, CVs and years of conversations that traditional databases struggle to understand together.
Belgian recruitment software startup Spott is building its business around making that accumulated context usable again. The company has raised a $21 million Series A led by Balderton Capital, with Base10 Partners, Y Combinator and Fortino participating. Combined with its earlier $3.2 million seed round, Spott has now raised $24.2 million. Co-founder Manu Vanderveeren said the new financing values the company at more than $100 million.
The funding headline is straightforward. The more interesting part is what Spott is trying to replace.
An ATS that remembers what recruiters actually learned
Traditional applicant tracking systems are good at storing structured fields: job titles, companies, CVs, stages and contact details. But recruiting decisions often depend on information that appears later in conversation: salary expectations, timing, willingness to relocate, reasons for leaving, client preferences or whether somebody quietly became open to a new role.
Spott combines an ATS and CRM with a vectorized, semantically searchable database. Its matching system can evaluate a candidate using not only a CV, but recruiter notes and recent interactions, then explain why the person matches a role. Recruiters can apply hard filters such as location, salary or availability before semantic ranking and retain control over the final decision.
That architecture changes the value of an old recruitment database. Instead of searching only for words that already appear in a profile, recruiters can describe what they need in natural language and search against the broader context accumulated around each candidate.
One company-hosted case study illustrates the difference. Executive-search firm H.W. Anderson had accumulated more than 100,000 candidate profiles, including years of call notes that were difficult to search in its previous system. After moving to Spott, those conversations became queryable alongside the rest of the candidate record.
The result Spott is pursuing is less “AI writes recruiter emails” and more a recruitment database that continuously turns conversations into usable memory.
The database updates while recruiters talk
Spott’s product now extends well beyond candidate search.
Its conversation layer can transcribe calls, generate summaries, suggest profile changes and create follow-up tasks from what was actually discussed. Its “Living Database” can propose updates when information changes, enrich missing email, phone or LinkedIn details, detect job movements and attach source reasoning to suggested changes.
Those records then feed other parts of the system: semantic matching, personalized outreach, client presentations and analytics.
Spott also combines email, WhatsApp, SMS, calendars and supported phone systems with multi-channel outreach, while its client-facing tools can reformat CVs, generate candidate reports and publish live candidate portals. Analytics track placements, fees, conversion rates and pipeline bottlenecks.
The important architectural idea is that each feature works from the same underlying context, rather than forcing agencies to stitch together a separate ATS, CRM, notetaker, outreach tool and reporting system.
Spott is also making the ATS readable by AI agents
One of the less-covered pieces of Spott’s strategy sits outside its own interface.
The company exposes an API and an official Model Context Protocol integration for Claude and ChatGPT, allowing external AI systems to interact with recruitment data stored in Spott. Its integrations already span communications, calendars, sourcing tools, job boards, VoIP platforms and services including DocuSign.
Paris-based recruiting firm Flaire offers an early example of what that could become. In a Spott case study, the firm describes using Claude through Spott’s MCP to build agents that examine candidate conversations, compare them with open roles and prepare potential matches for recruiter validation. The human recruiter still approves or rejects the result.
That distinction matters for where Spott says the new financing is going.
The company plans to develop more proactive, agentic capabilities that can surface potential vacancies, detect candidate career changes and suggest next actions without waiting for a recruiter to manually search for them. Spott says consequential actions remain under human control rather than automatically moving candidates or presenting them to clients.
As CEO Lander Degrève puts it, “Recruitment succeeds through trust and human judgement.”
The company’s product bet is that AI should increasingly handle the information work surrounding that judgment.
Why Balderton is funding the expansion now
Spott was founded in 2024 by Lander Degrève, Manu Vanderveeren and Samuel Smeys, university friends from KU Leuven who later worked at Bain, McKinsey and BCG. Their consulting work repeatedly exposed them to recruitment companies running core operations across ageing ATS products and disconnected software.
The company says revenue has increased more than tenfold since the beginning of 2026, while Vanderveeren says more than 500 recruitment agencies now use Spott. Balderton reports that roughly 20% of revenue is already generated in APAC and 15% in the U.S.; customers include H.W. Anderson and the 200-person CGP Group, which has deployed Spott across 12 countries.
That international pull explains part of the Series A.
Spott has opened operations in New York and Sydney, alongside Leuven, and plans to expand its engineering capacity and enterprise product as it grows from 44 employees toward roughly 60 by year-end.
Enterprise expansion creates a second requirement: recruitment databases contain unusually sensitive personal and commercial information. Spott says it is ISO 27001 certified and GDPR compliant, hosts data in the EU by default, encrypts information at rest and in transit, and does not use customer candidate or client data to train foundation models. Its disclosed subprocessors include OpenAI and Azure.
The Series A therefore finances more than geographic expansion. It is a test of whether an AI-native architecture can displace recruitment systems that have accumulated years of data but were designed primarily to store what recruiters knew, rather than reason across it.
Spott has already shown that agencies will migrate large databases into that model. The harder next test is whether its planned agents can become proactive enough to uncover commercial opportunities without crossing the line between useful automation and decisions recruiters still need to make themselves.
If it gets that boundary right, the future ATS may look less like a database recruiters maintain and more like an institutional memory that works alongside them.
