A family office can own the whole modern world and still struggle to answer a small question. What changed last month? The money may sit with private banks in Singapore and Zurich, in a New York fund, a Hong Kong brokerage account and a holding company that owns property. Add private equity, derivatives, art or wine, then scatter the records among online portals, spreadsheets and monthly PDFs. The total is valuable. The view of it can be surprisingly fragile.
Canopy makes software for that gap. The Singapore company collects positions, transactions and cash from custodian feeds, SWIFT messages, statements and private-market portals. It normalizes the data, checks that the pieces agree and keeps a trail back to the source. Only then does it draw the charts, calculate performance, produce reports or let an AI answer a question about the portfolio.
This order of operations is Canopy's argument about the wealthtech market: the interface is visible, but the plumbing is the product. The company says its system now covers more than $120 billion in assets under reporting for over 60 institutional clients in more than 20 countries. It connects to 250-plus custodians. Those figures describe scale, but they also describe maintenance. Bank formats change. PDFs arrive in seven languages. A security may be named three different ways. Every clean dashboard hides a small industrial process.
The answer before the dashboard
Canopy's origin predates the current AI enthusiasm. Tanmai Sharma, a former Deutsche Bank managing director, founded Mesitis in 2013 after seeing an oddity in private banking: a person could be sophisticated enough to hold accounts at several institutions and still have no accurate view across them. The first Canopy product launched in late 2014 as an account-aggregation and portfolio-visualization service. In June 2015, the company raised $3 million and said it already handled S$1.1 billion in assets under reporting.
The early wedge was practical. Canopy could accept data even when a bank did not provide a convenient API. A statement in PDF form was not a dead end; it was raw material. That mattered especially in Asia and Europe, where open banking for private wealth lagged the United States. By 2019, the company's then-chief technology officer, Amit Gupta, said its automation could process more than 1,500 kinds of bank statement.
“Data before AI.”Canopy's shortest product principle - and probably its most consequential
Today, the ingestion menu is broader: more than 100 direct data feeds, 19 SWIFT message types, PDF parsing and automated sourcing from private-equity and venture-capital portals. Canopy says the parser has 111 configurations and reads seven languages. If a bank remains unsupported, the company can build a parser and reuse it. The promise is bank-agnostic coverage, not dependence on the friendliest institutions.
One dataset, four jobs
Where the tedious work becomes the moat
Collecting data is only the first half of the job. Canopy matches transactions, positions and cash, then sends breaks to a review queue. It keeps a global security master so the same holding does not splinter into several identities. It fills certain gaps when banks omit opening or closing records for options, foreign-exchange forwards or accumulators. Manual entries are logged. A reported number can be traced back to a feed record, SWIFT message or statement line.
That lineage changes the kind of conversation a wealth manager can have. A principal does not merely see that net worth moved; the platform's long-running “Explainer” can separate the change into realized gains, unrealized gains, fund flows and currency revaluation. The user can drill toward transactions instead of opening five portals and reverse-engineering the month.
The same reconciled record supports daily revaluation, cost basis, profit and loss, accruals and corporate actions. It also feeds performance attribution, exposure and risk analysis, scenario modeling, benchmarks, cash-flow forecasts, mandate checks and capital-call schedules. Reports can be white-labeled, generated in batches and sent on a schedule. The data can leave through an API, a general-ledger export or a customer-built front end.
Read this chart as routes, not market share. Liquid securities typically arrive through bank and market feeds; private funds may require manager portals; property, art and wine often need controlled manual records. Canopy puts them into one model while preserving their different sources.
Who buys a quieter back office
Canopy sells to single- and multi-family offices, external asset managers, private banks, accounting firms and trustees. The common customer is not simply rich. It is operationally complex: many custodians, entities, currencies, beneficiaries and reporting audiences, with a lean team expected to produce institutional answers.
A single-family office wants one private view for the principal. A multi-family office needs repeatable reports without adding staff for every new relationship. An external asset manager wants bank-grade client reporting while remaining independent of any one custodian. Accountants need books that begin with reconciled investment activity; trustees need oversight across legal structures. Each buyer sees a different screen, but the calculation should come from the same underlying record.
The business model reflects the variety. Canopy is enterprise subscription software sold through a private demo, with pricing scoped to portfolios, custodians, users, reporting and service requirements. Customers can buy the software or combine it with onboarding, data upload and consistency-check services. This is SaaS with operational work attached - a sensible arrangement when the hardest edge cases are caused by other institutions' data.
The visible product is a report. The defensible product is the path back to the line item.
A crowded market, an Asia-first angle
Canopy competes with broad wealth platforms such as Addepar, Altoo, Asset Vantage, Eton Solutions and QPLIX, plus regional aggregators, specialist private-market tools and internal data teams. The stubborn incumbent is Excel. It is flexible, familiar and already installed. Its weakness appears when dozens of files must remain current, auditable and consistent across people.
Canopy's distinction is not a feature no rival could name. It is the combination: Asian custodian connectivity; direct feeds, SWIFT and multilingual statement parsing; public and private assets in one model; daily reconciliation; per-client data isolation; reporting, accounting and APIs on the same record. The company hosts data in Singapore, encrypts it in transit and at rest, runs each customer in an isolated database and holds ISO 27001:2022 certification.
That posture helped attract an unusually relevant investor. Credit Suisse embedded Canopy in a private-banking offering and bought a 10 percent stake in 2017. The partnership was both distribution and validation: a bank had backed software designed to show clients their assets across banks. Canopy says it now keeps customer names private as a rule. “Discretion is the product,” reads another company principle.
AI, but after the books balance
Mu Chen, an investor in Canopy, became chief executive in 2023 and led a product reset. In 2024, the company said it had removed confusing offerings and rebuilt its Visualizer around customizable dashboards and more than 100 metrics and attributes. In 2025 it widened private-asset coverage, strengthened allocation controls, added support for fixed coupon notes, launched an iOS app and published its first AI research paper. It also opened an office in Hong Kong.
The current platform includes natural-language portfolio chat, PDF analysis and machine-learning transaction classification. It can flag concentration, allocation drift, structured-product events and failing data feeds. The important qualification is that AI is opt-in. A client can use consolidation, reconciliation, analytics and reporting without it. Canopy says AI answers are grounded in reconciled positions and linked to sources.
This is a more credible pitch than dropping a chatbot onto a data swamp. It also creates a demanding standard. An assistant that gives a quick wrong answer about a movie is irritating; one that invents a portfolio exposure is dangerous. Canopy's opportunity is to make its accumulated reconciliation work useful at conversational speed without loosening the controls that made the data valuable.
Its 2026 direction pushes further into AI-assisted interpretation, alternative-asset connectivity and proactive workflows. Internally, the company says it pairs human leads with AI agents for repetitive research, drafting, monitoring and follow-up. Externally, it frames agents as assistants that surface issues and guide work rather than replace judgment.
The luxury of one honest number
Canopy fits between custodians and the decisions made above them. It is not a bank, asset manager or investment adviser. It is the translation and control layer: take inconsistent records, create a master version, then distribute it to the people and systems that need it.
That position can be unglamorous. It also becomes more useful as wealth grows more fragmented. Private markets produce documents instead of neat ticks. Families diversify across jurisdictions. A principal expects mobile access while an auditor expects lineage. The amount of data rises, but headcount does not need to rise with it.
The company began with a simple observation: having many banks is not the same as having a complete view. Its modern pitch adds AI, mobile software and 147 features, but the original problem remains intact. Canopy's best product is still a number that survived contact with every statement, currency and custodian behind it. In wealth management, that is less decorative than a dashboard - and more expensive to get wrong.