The AI reading finance's messiest documents - and turning them into data you can trust.
Smarter Alts Management · Founded 2013 · ~190 people
Every quarter, the back offices of the alternative investment world fill up with paper. Capital calls, quarterly statements, distribution notices, K-1 tax forms - each in its own format, each from a different fund, each requiring a human to open it, read it, and key the numbers into a system. Multiply that across hundreds of funds and thousands of positions, and you have one of finance's quietest, costliest bottlenecks.
Canoe Intelligence built its business on that bottleneck. The New York company uses artificial intelligence and machine learning to automate the collection, extraction and standardization of data from those documents. In practice, it means the PDFs that once demanded armies of analysts can be read, verified and delivered as clean data - ready to flow into whatever system an investor uses downstream.
The origin is telling. Canoe wasn't dreamed up in a lab; it grew out of an internal tool at fund-of-funds manager Portage Partners, built to tame the firm's own document workload before spinning out as an independent company. It is the kind of product that only gets built by people who lived the problem.
Today more than 325 institutional investors, capital allocators, wealth managers and asset servicers run on the platform, including names such as Blackstone and Hamilton Lane. In July 2024, Growth Equity at Goldman Sachs Alternatives led a $36 million Series C - a round that, by the company's account, tripled its valuation from the prior year.
"We eagerly anticipate driving the next phase of Canoe's evolution in partnership with Goldman Sachs."
For an allocator, the payoff is simple to describe and hard to achieve on your own: fewer hours re-keying data, fewer transcription errors, and something close to real-time visibility into a portfolio that used to update only when someone finished the paperwork. Canoe's modules automate document collection, extract the underlying figures, pull asset-level detail, manage tax documents, and deliver the results by API into clients' own systems.
The company processes more than 25 million documents a year. That scale is also the moat - every document teaches the system more about the strange, non-standard formats of private markets.
Because a general-purpose model can summarize a novel but struggles to reliably read a private equity capital call. Canoe's bet - one much of the market underweighted - was to train AI exclusively on alternative investments. In a domain where a misread number is a real financial error, narrow and accurate beats broad and impressive.
Automated collection of documents from portals, inboxes and fund administrators.
AI/ML engine that extracts and standardizes figures from unstructured alts documents.
Rapid asset-level intelligence, pulling underlying investment and portfolio-company data.
Streamlined tax document management, including K-1s and related filings.
Expert managed-service administration for clients who want Canoe to run operations.
Cross-document summaries that synthesize information across a client's document set.
Proprietary models trained exclusively on alternatives to power extraction and insight.
Roughly $62M raised across its known rounds, with the 2024 Series C reportedly tripling the valuation set a year earlier.
The technology that becomes Canoe is built inside fund-of-funds manager Portage Partners.
Canoe Intelligence is established as an independent business focused on alternative investment data.
Jason Eiswerth is appointed CEO; founder Seth Brotman continues as President and COO.
F-Prime Capital leads, with Eight Roads, to scale the platform.
Growth Equity at Goldman Sachs Alternatives leads, reportedly tripling the valuation.
Deeper investment in proprietary AI; client base past 325 institutions.
Alternative investing - private equity, credit, real assets, hedge funds - keeps growing, but its operational plumbing was built for a smaller world. Canoe sits in the data-infrastructure layer that allocators increasingly treat as essential rather than optional.
Allocators weigh Canoe against tools like Arch, Addepar, Solovis, Anduin and FundCount - and, most often, against the status quo of manual, in-house data entry. Canoe's edge is depth: AI and workflows built only for alts.
It uses AI and machine learning to automatically collect, extract and standardize data from alternative investment documents, delivering clean, verified data to investors and their downstream systems.
Institutional investors, capital allocators, wealth managers, family offices, fund-of-funds, pension funds, endowments and asset servicing firms - 325+ clients including Blackstone and Hamilton Lane.
Roughly $62M in total, including a $25M Series B in 2023 and a $36M Series C in 2024 led by Growth Equity at Goldman Sachs Alternatives.
New York City, at 307 Canal Street, with a team of around 190 employees.
Its AI is trained exclusively on alternative investment documents, prioritizing accuracy and verification for a domain where general-purpose models struggle.