13 / Investment case / AI FinTech
Project Meridian
A team investment case for an AI risk-intelligence company that sells to family offices. It sizes the market, picks a first buyer, chooses a fintech model over asset management, and sequences fundraising so the product earns credibility before it raises equity.
01 / The question
How does an AI company earn trust, and who pays for the wait?
The company builds AI risk intelligence for institutional wealth managers: downside-risk management and analytics that are hard to get outside a large hedge fund. Two decisions shape the case. Which clients should it sell to first, and in what order should it raise capital? The team's answer is family offices in Hong Kong first, run as a fintech that sells research and signals instead of a licensed asset manager, funded by non-dilutive money before any equity.
02 / Market research
Two markets, two jobs.
The research works outward from a wealth pool to a buyer and then to a channel. In Asia-Pacific it sizes the pool and the buyers. In the UK it looks at how software gets distributed.
| Figure | What it measures |
|---|---|
| 21%+ | Asia-Pacific private wealth as a share of the global total, projected to reach $99T by 2029. |
| $1.7T | Net cross-border inflows into Hong Kong and Singapore, 2019–2024. |
| ~4,000 | Family offices across Hong Kong and Singapore. |
| $5.8T | Intergenerational wealth transfer, which is driving family office growth. |
Hong Kong and Singapore are Asia's two main private wealth and booking centres, and the priorities of the wealth held there are capital preservation, downside risk management and alternatives. That is close to a description of the product.
| Factor | Finding |
|---|---|
| Market size | UK assets under management of $13T, 36% of Europe, concentrated in major wealth platforms. Platform-driven distribution allows one-to-many licensing. |
| Platform shift | A move toward platforms and model portfolios. More than 80% of advisors use model portfolios. |
| Channels | Wealth platforms and RIAs are the primary long-term license buyers. |
| Regulation | Regulatory transparency raises demand for differentiated analytics. |
The two markets do different work. Asia-Pacific is where the product is validated, through relationships. The UK is where it can scale, through licenses sold once to a platform and used by many advisors. Singapore follows Hong Kong as the natural next step for family offices, and the Cayman Islands appear only as a fund-structuring hub, not a place to win clients.
03 / The first buyer
Why family offices, and why Hong Kong first.
Family offices are small teams holding large, complicated portfolios, and their preferences shape the product as much as the price does.
| Theme | What the research found |
|---|---|
| Structural demand | Lean investment teams of 2–10 professionals manage complex, multi-asset portfolios. About 70% of Asia-Pacific family offices rely on customised mandates, not standard products. They want transparent, explainable, human-in-the-loop AI and have little tolerance for black-box or fully automated allocation. |
| Strategic fit | A relationship-led ecosystem rewards trusted, credible entrants. That means a lower customer acquisition cost, a faster time to revenue, and a higher chance of near-term monetisation than reallocating effort to new segments. |
The demand facts point straight at the design. A buyer that wants explanations and control will not accept a system that trades for it, so the product is built to inform a human decision, and every signal can be explained and audited.
04 / The product
An air-defence system for a family office.
The pitch is “generational preservation”. Independent family offices lack the risk infrastructure of large hedge funds. They have no internal “air-defence system” for regime shifts, correlation spikes and drawdowns, and building one in-house is costly, slow and operationally complex. The platform supplies hedge-fund-grade analytics without the hedge-fund overhead, and it enhances existing strategies without replacing human judgment.
From portfolio data to a decision
- Ingestlive portfolio data reveals hidden factor and correlation exposures
- ScanAI monitors global market regimes for early stress signals
- Signaltimely alerts to shift toward cash or defensive assets
- Family office decidesno recommendation, execution or portfolio control
05 / Competition and precedents
Priced below the incumbents, financed like the precedents.
The team compared the company with a global incumbent and a regional rival, then proposed pricing that undercuts both on cost and on flexibility.
| Player | Profile | Strength and limit |
|---|---|---|
| Bloomberg (large competitor) | Revenue of about $15B at a roughly 40% EBITDA margin. About $28k per user per year, and a typical family office pays $56–140k a year. | Unmatched data breadth and workflow integration, but generic analytics with no mandate-specific strategy improvement. |
| AQUMON (regional, Hong Kong) | QuantWealth platform for asset allocation and implementation. 2024 revenue of about $67.2M, earned through AUM-linked fees and institutional partnerships. | Scalable, regulated implementation, but pre-defined models with limited flexibility for bespoke mandates. |
| Proposed pricing | A tiered annual subscription per family office, plus a one-time implementation and customisation fee. | Cheaper than the competitors and more flexible than AUM-based platforms. |
An appendix table then follows four precedent intelligence and wealth-technology companies to see how they were financed and how they scaled.
| Company | Financing path | Outcome |
|---|---|---|
| SimCorp | Organic growth, IPO in 2000, strategic acquisition in 2023. | Scaled to global mission-critical infrastructure without managing assets. |
| Envestnet | VC-backed, IPO in 2010, PE-led take-private in 2024. | Scaled through workflow lock-in and distribution control, not discretionary management. |
| AlphaSense | VC-led from seed to late-stage growth equity, 2019–2024. | Scaled pure investment intelligence on per-seat subscription revenue and platform depth. |
| Yewno | Early-stage VC only, then stalled growth and inactivity. | Academic prestige and index partnerships, but no sustained institutional adoption. |
Reading the table, three of the four scaled without managing money, and the one that stalled had credibility without adoption. That is the argument for the sequencing later in the case: institutions have to use the product before capital can accelerate it.
06 / The model choice
A fintech first, an asset manager only if it earns the right.
The first phase is research as a service: non-discretionary signals, analytics and portfolio diagnostics that can be sold immediately, without licensing friction, to clients anywhere. A licensed asset-management business is held back as a second phase, once a track record and credibility are proven, to capture asset-based fees if the team wants them. The cost comparison explains why.
| Cost line | Annual cost |
|---|---|
| Hong Kong SFC licensing and setup | $60k |
| Compliance officer | $80k |
| Regulatory reporting and audits | $40k |
| Professional indemnity insurance | $20k |
| Fund administration, operations and reporting | $80k |
| Legal and risk support | $30k |
| Asset and wealth management, total | ~$310k, plus seed capital |
| FinTech path: one additional data compliance officer | ~$80k |
The six lines add up to the $310k. The deck's conclusion is that the fintech path is about 4x cheaper structurally and 16x cheaper in total. The 16x only works if the seed capital for asset management is roughly $1M, which the slide's “>$1 investment” appears to mean.
07 / Regulation
Stay a publisher, not an adviser.
The fintech model only works if it stays clear of advisory classification. The regulatory design is deliberately narrow.
| Layer | What it does |
|---|---|
| Core model | B2B only, for corporate and professional investors such as family offices. It delivers market intelligence, benchmarks and probability signals, and no recommendations, execution or portfolio control. Access is by a public “Apply for Access” platform, which establishes publisher status and avoids advisory classification. |
| Governance | A data compliance officer (about $80k a year) covering data integrity, bias, privacy and auditability, with no investment decision authority. A “snapshot rule” delivers outputs as periodic briefings, not live execution feeds, which reinforces the classification as a research publisher and not a trading venue. |
| Liability protection | An as-is data disclaimer, with analytics for benchmarking only. An anti-collusion covenant against coordinating trades between clients. A corporate gate that keeps consumers out and so avoids consumer-protection exposure. |
The deck's reference list includes the Hong Kong Securities and Futures Commission's code of conduct, Singapore's Financial Advisers Act, the MAS principles on fairness and transparency in AI, and the U.S. case Lowe v. SEC.
08 / The capital plan
Credibility before capital.
Each phase of funding is matched to what the company has to prove at that point.
| Phase | Capital and use | Outcome |
|---|---|---|
| 0–6 months | Non-dilutive funding: the Cyberport Incubation Program (up to HK$500K), the Enterprise Support Scheme (up to HK$10M), the British Business Bank (up to £10M) and UK Research and Innovation (fund capacity £100M). Prioritises institutional trust over valuation. | Embedded institutional usage and proof of relevance, with no equity dilution. |
| 6–24 months | A strategic minority equity round from bank-affiliated venture arms, asset-manager innovation funds and wealth-tech corporate VCs. Spent on platform integrations with wealth-tech infrastructure and selective go-to-market, not headcount. The product is already validated, so equity accelerates reach, not experimentation. | Structural scale without bloated headcount, and faster access to family offices through distribution partners. |
| 24+ months | A choice of two paths, below. The objective is to preserve strategic choice while scaling globally. | See paths A and B. |
| Path | Funding and valuation | Outcome |
|---|---|---|
| A · FinTech-only scale | A growth equity round from UK and Asia-Pacific fintech VCs, late-stage fintech VCs or corporate venture arms. Valuation is benchmarked to listed fintech SaaS comparables. | A high-margin, asset-light intelligence platform, with international expansion, deeper AI and broader integrations. |
| B · Optional asset-product extension | Strategic minority equity, if pursued. ETFs, separately managed accounts or model portfolios through partnerships with mid-sized, innovation-led asset managers. | Asset products as an extension of proven intelligence, available only once recurring revenue and family-office trust are established. |
09 / The ask and its limits
Strategic equity, with control and choice intact.
The ask is strategic equity from family-office-aligned AI and fintech VCs, to compress the time to scale while preserving founder control and validating product-market fit. The economics the team points to are those of software: distribution through several channels (wealth managers, ETF issuers and platforms), license-based scaling with signal volume, high margins on light capital, and recurring revenue that supports software valuation multiples. The conclusion is that staying fintech-only and family-office focused in the early phases removes the main barriers to adoption, and that sequencing capital this way turns credibility into scale without giving up flexibility or ownership.