The archive · AI & Models · Product decision · 2024–2026
Capsa AI's private-capital knowledge bet: $18M on 14x ARR
Capsa AI bets funds will pay for an AI knowledge layer over emails, CRMs and PDFs; 14x ARR growth, 100% renewal and an $18M Series A.
Capsa AI
What the business is
Builds an 'AI operating system' for private capital: indexes a fund's CRM, email, SharePoint and proprietary data into one searchable knowledge layer used across sourcing, due diligence, portfolio monitoring and back office.
Starting capital：£1.7M seed eight months after founding (September 2024), then an $18M Series A co-led by TX Ventures and Pivot Investment Partners with Bek Ventures, Antler, Outward VC and Cornerstone VC (June 2026) — $20M total.
How it started
Danyal Oezduezenciler spent more than seven years in private equity and investment banking at AEA Investors, Citi and Deutsche Bank, sitting through over 50 due diligence processes, before founding Capsa AI in 2024 with Callum Downie, whose background spans AI and machine-learning products at ThirdEye Labs and Standard AI. Their observation: private capital manages more than $15 trillion in assets yet runs on emails, PDFs and spreadsheets, with professionals losing an estimated 500+ hours a year to manual search at an industry cost of roughly $35 billion — while data volumes double every three years and headcount does not.
What happened
Capsa raised a £1.7M seed in September 2024, eight months after founding, then scaled through founder-led sales: it opened a New York office, moved Oezduezenciler to the US to lead expansion, and reported 14x year-on-year ARR growth with no paid customer acquisition, a 100% renewal rate, net dollar retention above 122% and more than half of booked ARR on multiyear contracts. The platform indexes millions of internal documents and conversations at several large private capital firms. The $18M Series A, co-led by TX Ventures and Pivot Investment Partners with Bek Ventures and existing investors Antler, Outward VC and Cornerstone VC, funds US expansion, engineering and go-to-market hiring and more agentic capabilities that execute workflows from the firm's own knowledge.
No ending yet — it is still running.
Background
Capsa AI, founded in 2024 by Danyal Oezduezenciler and Callum Downie, is betting that private capital will pay for a vertical AI layer over its own institutional knowledge. Oezduezenciler, who spent more than seven years at AEA Investors, Citi and Deutsche Bank and sat through over 50 due diligence processes, saw the industry's paradox firsthand: funds manage more than $15 trillion in assets, yet critical information lives in emails, PDFs, CRMs and spreadsheets that teams must search manually.
The product is an 'AI operating system for private capital': it indexes a fund's CRM, Outlook, SharePoint and proprietary data into a unified knowledge layer mirroring how the firm evaluates deals — sourcing, due diligence, portfolio monitoring and back office. Instead of a standalone assistant, Capsa embeds into existing workflows, makes institutional knowledge searchable and is adding agentic capabilities that complete multistep tasks from a firm's own records. It is SOC 2 Type II certified, offers single-tenant deployment and regional data hosting, and does not train on customer data.
Capsa raised a £1.7M seed in September 2024, eight months after founding, then grew through founder-led sales with no paid customer acquisition. By June 2026 it reported 14x year-on-year ARR growth, a 100% renewal rate, net dollar retention above 122%, more than half of booked ARR on multiyear contracts, roughly 70% weekly and 40% daily active usage, and deployment at several of the world's largest private capital firms, where it indexes millions of internal documents and records.
The $18M Series A, announced June 10, 2026, was co-led by TX Ventures and Pivot Investment Partners with Bek Ventures and existing investors Antler, Outward VC and Cornerstone VC, bringing total funding to $20M and funding US expansion and agentic product work. The bet's edge is domain specificity: the founders argue generic AI copilots cannot encode how a particular fund thinks, while Capsa's model-agnostic, security-first knowledge layer compounds with every deal it indexes.
What has to be true
- Private capital is one of the most data-intensive industries on earth yet runs on fragmented documents, emails and spreadsheets.
- An estimated 500+ hours per professional per year go to manual search and analysis — a $35 billion annual cost the industry can price.
- Indexing a fund's own history encodes how that firm evaluates deals, which generic copilots cannot do.
- Security posture — SOC 2 Type II, single-tenant deployment, no training on client data — clears the compliance bar major funds require.
- 14x ARR growth, 100% renewal and >122% net dollar retention show enterprise demand, not just investor narrative.
What can be applied
Vertical beats horizontal in knowledge work: encode how a fund actually operates, win on security and embed into workflows — then hard metrics (14x ARR, 100% renewal, >122% NDR) do the fundraising.
Aftermath
As of June 14, 2026, Capsa AI operates from London and New York with 14 employees, serving several of the world's largest private capital firms across buyout, growth equity and private credit, with further enterprise rollouts underway in the US and Europe. The $18M Series A brings total funding to $20M and is earmarked for US expansion, engineering and go-to-market hiring, and agentic AI that executes workflows from a firm's indexed knowledge. Revenue figures beyond growth rates had not been disclosed.
Sources
- Capsa AI raises $18M to expand its AI platform for private capital
- Capsa AI Raises $18 Million Series A To Build AI Operating System For Private Capital Firms
- Capsa AI raises USD 18 mln Series A for private capital AI platform
spotted an error? The archive wants to know.
Your turn
You just read one. Describe what you are building, and see who is betting on the same thing.
Free account · 3 free questions · no card