Investors

The case, with its sources.

Market and competitive analysis behind Phiuto, summarised from our research base. Every figure below names its source; ranges are shown where sources disagree; our own constructions are labelled as such. The full memoranda — with links, verification tiers and disclosed conflicts — are available on request.

Early stage · in active developmentWedge locked: procurement · Sweden firstNo paying customers yet — claims are hypotheses unless sourced

01 — The stakes

The spend is already budgeted.

These are the stakes our answers decide — actual spend flowing through procurement and external data. They are not our revenue.

€2T / yrEU public procurement — ~14% of EU GDP, across 250,000+ public authoritiesEuropean Commission · verified at source

SEK 800B+ / yrSwedish public procurement — roughly one fifth of GDP; ~4,000 procuring organisationsUpphandlingsmyndigheten · Konkurrensverket

$49.2B 2025Global financial market data spend, +6.5% YoY — a record every year for a decadeBurton-Taylor · verified at source

Weeksof analyst work per sourcing decision today — ~18,500 procurements advertised per year in Sweden aloneThe wedge problem

02 — Market

Two cases. Defended in that order.

Case 1 — the core engine and the procurement wedge — is what we defend in diligence. Case 2 — the full platform across five applications — is why it's venture-scale. The beachhead is identical in both.

Case 1 — engine + wedgeCase 2 — full platform

TAM today (2025)~$11B — procurement software + analytics slice, de-duplicated~$30–35B — adds company-data intelligence and the external-data slice of integration

TAM 2030–31~$23–30B at ~9.5–11% category CAGR~$60–80B; the agentic-AI layer alone forecast at 40–46% CAGR

SAM — Europe (2025)€2–4B — supplier discovery, qualification and company data€7–9B core categories, → €16–20B by 2031

Vision arena€2T/yr EU public procurement — the spend our answers decide$350–400B/yr external-data economy → $550–750B early 2030s. The arena, not addressable revenue — most of it is consumer data today

SOM — Nordics, 36 moIdentical in both cases: ~€2.5M ARR base case — ~100 organisations from a ~10,000-organisation Nordic universe at ~1% penetration, €25k ACV. Bottom-up, every assumption labelled; range €0.8–6M. The wedge is the entry either way — Case 2 raises the ceiling, not the beachhead.

TAM

~$30B → $60–80B by 2030

SAM

€7–9B Europe · wedge €2–3B

SOM

~€2.5M ARR · Nordics · 36 mo

Sources: Apps Run The World, Mordor, Fortune BI, Grand View, MarketsandMarkets, Burton-Taylor, Market Data Forecast. Research-firm sizings disagree — we present ranges and name the spread; sums and de-duplications are our own construction, labelled in the memorandum.

03 — The long-term customer

Agent demand is measured, not forecast.

The eventual customer is the AI agent — and its consumption of external data is already visible in infrastructure traffic and enterprise spend.

15×growth in AI "user action" crawling in 2025 — agents fetching pages to complete a user's task. Cloudflare Radar, measured

$37B · 3×enterprise generative-AI spend in 2025, tripled YoY — and the layer connecting models to external systems and data is the least-built part: $1.5B of $18B infrastructure spend. Small today means early, not absent. Menlo Ventures

40% by 2026of enterprise applications will embed task-specific AI agents, up from under 5% in 2025; a third of enterprise software agentic by 2028. Gartner

HTTP 402the web is starting to meter and charge agents for access (pay-per-crawl). Free browsing is ending — which converts scraping into exactly our problem: strategy across licensed, priced, heterogeneous sources with cost accounting and provenance. Cloudflare, secondary

The counterweight, said before you ask

Gartner also predicts over 40% of agentic-AI projects will be cancelled by end-2027. Read correctly, that supports the thesis: agent applications will churn; the layer they all depend on — reliable, verified external data — is where durable value sits. Our plan is funded by the procurement wedge; the agent layer is upside timing, not plan-critical revenue.

04 — Competitive landscape

Heavy capital at both ends. Nobody owns the objective.

Procurement suites are buying supplier-discovery capabilities at the top; web-data APIs for AI agents are raising at unicorn valuations at the bottom. In between, no company's unit of execution is the completed, verified goal.

CategoryWhat they sellWhat none of them do

Supplier discovery & dataScoutbee→Coupa · Tealbook+Supplier.io · VeridionA pre-built supplier database with a search boxConstruct an acquisition strategy per objective; measure completion; report gaps

Nordic procurement techSievo · Ignite · Kodiak Hub · TendiumAnalytics on your own spend; SRM on existing suppliers; tender workflowsOutside-in discovery and qualification of new suppliers, with evidence

Nordic company dataVainu · Roaring · D&B/BisnodeRegistry-sourced firmographics for sales and CRMQualification against technical and certification criteria; verification; per-field provenance

Agent-native web dataParallel · Exa · Bright Data · ApifySearch, extraction and proxy APIs — the means of acquisitionOwn the objective: completion criteria, source strategy, cross-source verification, licensing per source

Incumbent suitesCoupa · SAP Ariba · JAGGAER · IvaluaEnd-to-end spend management for global enterprisesServe the mid-market; work outside their own network's transaction data

$150M+of VC validated buyer-side supplier discovery — then both leaders were absorbed within ~8 months (Coupa/Scoutbee, Tealbook/Supplier.io). The category earns exits; the independent field is momentarily empty.

$340M+ · $2Braised in ~a year for agent-native web data (Parallel, Exa), at valuations to $2B. "AI agents need a data layer" is Sequoia's and Benchmark's priced position, not our speculation.

Unfundedthe middle of the stack: between the database companies and the retrieval APIs, nobody owns completion criteria, verification, provenance — and the honesty to deliver 43-of-50 with a gap report.

Our capability comparison is design intent versus their shipped traction — we say so. Full profiles, funding map with disclosed tracker conflicts, capability gap matrix and threat postures are in the Competitive Landscape Memorandum.

05 — Where we are

What we're proving now.

We're an early-stage startup building Phase 0: one procurement goal type, end to end, in Sweden. These are the numbers that decide everything — instrumented in the product from run one.

demoGoal completion rate — finished without human rescue

businessCost per completed goal — must trend down

demoAccuracy — % of material fields correct on human audit

businessHuman intervention rate — must trend down

demoCoverage — % of the requested universe found

businessSource reuse rate — must trend up: the flywheel, measured

The demo metrics prove the system works. The business metrics prove it's a software business — if cost per goal doesn't fall with volume, we'd rather know in month six.

The investor pack

12-slide deck · Market Opportunity Memorandum · Competitive Landscape Memorandum — with sources, verification tiers and disclosed conflicts. We're not formally raising; we talk early to investors who know this space.

Request the pack