Espoo, Finland

Every fund on the shelf, backed by evidence.

Fund Anvil gives fund-shelf and product-governance teams at banks and fund platforms cited evidence on every fund they offer, from each fund’s own public documents.

Raising €1M in H1 2027.

Claravium Technologies

The shelf

A bank or fund platform offers hundreds of funds.

Most come from outside managers. Each fund carries a prospectus, a KID, an SFDR annex and periodic reports.

  • 100s funds on a typical shelf
  • 1,000s documents behind them

Every distributor’s problem

Distributors must stand behind every fund they offer.

Today their own criteria (exclusions, sustainability minimums, naming rules) are checked against fund documents by hand.

Checking fund documents is manual today, so it is hard to show from evidence why each fund is on the shelf.

The gap today

The hidden problem

The documents themselves can’t be trusted.

It surfaces only at marketing review, in distributor due diligence, or when the regulator asks.

Stale

KIDs and fund pages go out of date and keep old costs and risk scores.

Contradictory

Costs, risk scores and holdings differ between prospectus, KID, SFDR annex and reports.

Why now

Distributors must keep reviewing every fund they offer.

MiFID II product governance and ESMA’s 2023 guidelines require regular checks that each fund still fits its target market. Today that review is largely manual.

Sources: MiFID II, Directive 2014/65/EU, Art. 16(3) and 24(2); ESMA Guidelines on MiFID II product governance requirements (ESMA35-43-3448, 2023).

Every fund. Every review cycle.

MiFID II · ESMA 2023

The product

Fund Anvil: cited evidence on every fund.

  • Reads every fund’s full public document set.
  • Checks it against the distributor’s own criteria.
  • Flags stale and contradictory documents.
  • Cites the page behind every finding.
  • In the app, by API, or through MCP in your own AI tools.

Fund Examination covers return, risk, cost, flows, holdings as published, gaps and limits. A continuous monitor for funds on a distributor’s shelf, and for managers who want their own published record inspected privately.

Disclosure Review compares the documents against each other and flags each inconsistency with a severity and the effort to resolve it. An evidence layer beside counsel, not in place of it.

Fund Examination

What one examination covers.

Six sections, with every fund compared side by side. Figures from a test on one fund manager’s public documents.

Return

Each fund against cash and against its own benchmark, year by year.

Risk

One fund: deepest loss −56%, worst calendar year −53%, 23 months to recover.

Cost

One fund: a 1.15% ongoing charge, but 2.42% actually charged in 2025.

Money leaving

Four funds lost 28% of value to redemptions in a year they returned +3% to +32%.

Holdings

Not one individual holding named in any audited annual report.

What is published

Which figures each fund publishes, and which none of its documents state.

Evidence from a public-document test

Tested on one fund manager’s public documents.

A full Fund Examination and a Disclosure Review, from public files only.

Each finding comes with a severity and a fix.

  • 40 public documents read
  • 4 funds examined
  • 11 findings, each cited
Example from a test on public fund documents
Fund A property fund
Finding Fund page advertises an 8–10% a year return target; the KID forecasts 4.9% a year after costs; 8% and 10% are the performance-fee hurdles in the fund rules.
Severity High
Effort One page edit
Evidence Fund page; KID p.2; fund rules §11

Alternatives

Data and process are covered. Evidence is not.

Fund data hubs

FE fundinfo and Morningstar store and distribute documents and data, not per-finding evidence.

Product-governance workflow

Ruleguard and Deloitte OneView organise reviews. They track the process, not what is inside the documents.

Ratings and general AI

Ratings terminals score funds. General AI summarises without findings you can cite.

Manual review

Spreadsheets are slow, uneven and hard to evidence.

The cited consistency layer between those tools: findings anchored to the issuer’s own files, with severity and effort to resolve, as a continuous monitor.

Evidence, not advice.

Market

About €160M a year in Europe, starting in the Nordics.

Buyers are product-governance and fund-shelf teams at banks, fund platforms and wealth managers in Europe, including the UK and Switzerland, plus fund managers and ManCos who want a private continuous monitor of their published record.

Bottom-up Europe TAM is about €160M a year (base; low about €50M, high about €430M): about 200 large shelf owners, about 1,500 mid-sized and about 4,000 smaller firms, plus about 2,000 manufacturers.

Nordics first, then DACH and Benelux, then the wider EU, selectively beyond. First goal: 3–5 design partners across shelf owners and managers. Demand is driven by MiFID II and ESMA product-governance duties to review each fund on the shelf for as long as it is sold.

Claravium estimates. TAM, SAM and SOM are annual figures.

  • ~€40M SAM · Nordics, DACH and Benelux
  • ~€10M Nordic beachhead
  • €3–4M Year-5 SOM · ARR, base case

What it unlocks

Distributor side

  • Knows what it offers and why.
  • Knows which managers to chase.
  • Advisers and fund pages get clear, cited facts.

Fund-manager side

  • Fixes its documents before distributor due diligence.
  • Runs a private continuous monitor of its own published record.

An evidence layer beside counsel, not in place of it. Not audit, not compliance advice, not investment recommendations.

A continuous monitor, not a one-off report.

Subscription / SaaS.

Banks, platforms and wealth managers

The lead offer is a continuous monitor on retainer for the funds on a distributor’s shelf.

Fund managers and ManCos

A private continuous monitor of their own published record, before distributor due diligence or document relaunches.

The path is a paid pilot, then a cohort of 3–5 design partners. Pricing is shared in conversation.

Roadmap

From Nordic pilots to European expansion.

Nordics first, then DACH and Benelux.

2026 Q4 – 2027 H1

Pilots with Nordic distributors and a €1M raise.

2027 H2

First paying Nordic customers.

2028

Expansion to DACH and Benelux.

The ask

Raising €1M in H1 2027.

Angel / pre-seed.

  • Product: build out Fund Anvil and Disclosure Review.
  • Pilots: Nordic distributor pilots through to first paying customers.
  • Team: engineering and go-to-market, ready for DACH and Benelux by 2028.
  • €1M H1 2027 · angel / pre-seed

Access

In the app, by API, or through MCP in your own AI tools.

  • App
  • API
  • MCP

Founder, Claravium Technologies

Tuomas Siren

Founder-led, with 30+ years in software architecture and infrastructure security across financial services, telecom and technology. That engineering discipline goes into Fund Anvil: every finding cites page and line in the issuer’s own public documents, outputs are versioned and reproducible, and the claim is strictly bounded — evidence, not audit, compliance advice or investment recommendations.

Focus: deepen the pain, refine the offering, sign 3–5 design partners across shelf owners and fund managers, then expand the team.

At Slush: meeting VCs, and banks, fund platforms and fund managers as prospective design partners.

Espoo, Finland

tuomas.siren@claraviumtechnologies.com

linkedin.com/in/tuomassiren