Updated:
Squirro
Dorian Selz built Squirro from a Zurich semantic search lab into an augmented intelligence platform used by the ECB, Bank of England and ING.
Squirro
Dorian Selz and Toni Birrer launched Squirro in 2012, commercializing a decade of semantic search research from the University of Zurich. The firm grew out of a precise observation: enterprise search had stalled, and the next wave required insights pulled from unstructured data across siloed systems. Early backing came from Swiss and European venture investors, positioning Squirro not as a horizontal SaaS tool but as a verticalized AI layer for financial services, insurance, and government. The founding team retained technical leadership through each funding round, an unusual continuity in an acquisition-heavy European AI market. Squirro operates as a retrieval-augmented generation platform, combining large language models with proprietary knowledge graphs and a semantic search backbone. The stack ingests structured and unstructured data — emails, research reports, regulatory filings, internal wikis — and surfaces actionable answers with citation trails. Deployment typically spans risk analysis, client intelligence, and compliance workflows. Named customers include the European Central Bank, Bank of England, and ING. The firm also maintains an out-of-the-box solution for asset management research, pulling in sell-side reports, news feeds, and portfolio data to generate analyst-ready briefs. Geographically, Squirro serves clients across Europe, with growing clusters in North America and Singapore. Headquartered in Zurich, Squirro has added offices in London, New York, and Singapore to support its institutional client base. While headcount and financial metrics remain private, the firm's client list signals a concentrated enterprise-governments niche rather than a volume-growth strategy. In 2023 Squirro integrated GPT models into its core platform, coupling external LLM calls with its own security and access-control layer — a move that allowed regulated institutions to use generative AI without data leaving their perimeter. The firm has also released sector-specific application templates for wealth management and insurance underwriting. Squirro's structural differentiator is its head start on citation-backed AI inside compliance-heavy environments. While most enterprise AI vendors retrofit large language models onto generic productivity suites, Squirro built its retrieval engine during the pre-transformer era and adapted it outward when GPT models arrived. This means every generated insight carries a traceable source path, a feature that banking and insurance regulators now demand. The firm remains founder-led, privately held, and Swiss-headquartered — operating as one of the few independent European AI platforms competing against hyperscaler ecosystems on data-residency and auditability grounds.
General information
Firm type
Asset Manager
Year founded
2012
AUM
Undisclosed
Location
Region
Europe
Country
Switzerland
City
Zurich
Corporate office
Zurich, Switzerland
Additional offices
London · New York · Singapore
Principals
Dorian Selz
Co-Founder & CEO
Toni Birrer
Co-Founder
Sector focus
Frequently asked questions
Who runs investment decisions at Squirro?
Squirro is a privately held enterprise software company, not a fund. Dorian Selz and Toni Birrer retain operational and strategic control as co-founders, with Selz serving as CEO. Equity funding rounds have been led by European venture investors, but the firm does not manage third-party capital as an asset manager or family office. Investment decisions — meaning capital allocation within the company — rest with the executive team and board.
How does Squirro's augmented intelligence differ from standard enterprise search?
Squirro combines retrieval-augmented generation with a proprietary knowledge graph and semantic search layer. Standard enterprise search indexes keywords; Squirro connects unstructured data across emails, regulatory filings, and research reports, then surfaces answers with traceable citation trails. The platform predates the current wave of GPT wrappers — its indexing engine was built in the early 2010s — and added large language model integration in 2023 without losing the audit-trail feature that regulated institutions require.
Which regulated institutions use Squirro, and for what purpose?
Named clients include the European Central Bank, the Bank of England, and ING (per the firm's official communications). Typical use cases center on risk analysis, regulatory compliance monitoring, and client intelligence — pulling fragmented internal data into a single query layer that produces sourced answers. The platform is deployed on-premise or in private cloud environments so that data never leaves the institution's perimeter, a requirement for central banking and insurance regulators.
Is Squirro structured as a single-family office or does it operate more like a venture firm?
Neither. Squirro is an enterprise software company that sells augmented-intelligence platforms to financial institutions and governments. It is founder-led and venture-backed, not a family office or a fund. The firm does not manage external capital, make direct investments, or operate a venture arm. It generates revenue through platform licensing and professional services.
What is Squirro's known posture on co-investments alongside external GPs?
Squirro is not an investment firm and does not participate in co-investments alongside general partners. It does not deploy capital into portfolio companies, funds, or direct deals. The firm's only relationship with the investment community is as a technology vendor — selling its augmented intelligence platform to asset managers, banks, and insurance companies for their internal research and compliance workflows.
Where does Squirro's technology originate?
The core technology emerged from semantic search research conducted at the University of Zurich. Co-founders Dorian Selz and Toni Birrer commercialized that research starting in 2012, building a platform that could synthesize insights from unstructured enterprise data well before the large-language-model boom. That decade-plus head start on retrieval and citation infrastructure remains the firm's technical moat, particularly inside compliance-heavy environments that cannot tolerate the hallucination risk of standalone large language models.
Which sectors does Squirro explicitly target?
Squirro focuses on financial services, insurance, and government — three sectors where unstructured data is abundant and regulatory pressure demands auditable AI outputs. The firm also serves asset management and wealth management clients with pre-configured application templates. It does not market heavily to consumer technology, healthcare, or manufacturing verticals, though its general-purpose semantic engine can be applied more broadly.
Profile maintained by Altss using OSINT (open-source intelligence), regulatory filings, licensed data partners, and verified direct submissions. Read the methodology. Last updated: . Continuous refresh with full update cycles at least every 30 days.
Need institutional-grade insight on family offices?
Altss delivers:
Prefer a guided tour?
We’ll walk you through: