The orchestration layer
between people & their stack.

ANi is the AI orchestration layer mid-market enterprises need to safely deploy frontier AI across their stack. We are not a foundation model company. We run on Anthropic, Google and Microsoft models and add the layer enterprises actually buy: universal integration, role-aware workflows, governance, and a contractually-enforced no-lock-in commitment.

Our beachhead is mid-market financial services in EMEA — 250–2,500 employee firms in asset management, insurance, fintech and regional banking. The sector has acute licence bloat, regulated complexity, and a procurement culture that has been burnt by incumbent vendor lock-in. ANi's no-lock-in stance is not marketing copy — it is a paid handover SLA built into every contract.

Pricing is seat-based with token overage and optional outcome-based add-ons. Predictable for CFOs, transparent for procurement, and economically sound for ANi because token costs pass through with protected margins.

The long-term vision — any size, any industry, any location — is preserved. The funded version of the plan executes a 24-month wedge to earn the right to it.

From "AI Engine" to "Orchestration Layer"

Foundation models cost hundreds of millions to train and billions to keep current. Claiming a proprietary engine we cannot fund eroded credibility. ANi is honestly positioned as an orchestration, integration and governance layer on top of frontier models. Our IP is in routing, memory, role-aware context, governance, and the implementation methodology — not in the underlying weights.

From Horizontal Manifesto to Beachhead Wedge

"Any size, any industry, any location" is a vision, not a go-to-market plan. The funded version of ANi wins decisively in mid-market financial services in EMEA in the first 24 months, then expands geographically before going cross-vertical.

From Hardware-Inclusive MVP to Software-First

The secure dumb-terminal vision is preserved as a phase-two product, post Series A. Including it in the MVP doubles the funding ask, adds a second go-to-market motion, and creates supply-chain risk with no offsetting commercial validation. The software platform comes first.

From Token Billing to CFO-Friendly Pricing

Pure token billing is how infrastructure providers price — and CFOs hate it. ANi prices by seat with a generous included token allowance, transparent capped overages, and an optional outcome-based add-on for sophisticated buyers.

"Better AI" is not a moat in 2026. ANi commits to three:

The trust moat is the marketing wedge. Integration depth is the technical wedge. Methodology is the delivery wedge. They reinforce each other.

The reasonable investor question is: why won't Microsoft Copilot or Google's Gemini Enterprise eat this category? Three structural reasons:

• Hyperscalers are bad at mid-market. Their sales motion, pricing, and support model are built for enterprise. Mid-market firms do not get named account managers, custom integration, or a methodology team. ANi is built for them.
• Hyperscalers are bad at vendor-neutral integration. Microsoft Copilot is structurally biased toward Microsoft's stack. Google Gemini Enterprise is biased toward Google's. ANi runs across all three frontier model providers and integrates whatever the customer already owns.
• Hyperscalers are bad at services-led implementation. They sell software; they don't run change programmes. Mid-market customers need both. ANi packages the methodology, the program management, and the documentation as part of the offering — what mid-market firms used to pay Big Four for.
• Hyperscalers cannot credibly offer no-lock-in. It is structurally incompatible with their commercial model. ANi can — and that builds permanent trust.

This is not a claim that hyperscalers are weak — they aren't. It is a claim that the three weaknesses they all share define exactly the segment ANi addresses.

Overage: any token consumption beyond the included allowance is billed transparently with caps, alerts, and finance-team controls. Overage cost is never a surprise.

Outcome-based add-on: for sophisticated buyers, ANi offers gain-share contracts tied to documented hours saved or software licences retired. This is opt-in and reserved for customers who want to share upside.

Implementation: fixed-fee engagement using the ANi methodology. No T&M surprises. Scope is defined; overruns are ANi's risk.

Handover package: paid, SLA-backed exit assistance — sold at contract signing. This is the no-lock-in commitment made commercial.

Every iPaaS vendor since 2015 has promised "integrate with anything." ANi's integration claim is structurally different: instead of pre-built connectors for known APIs, ANi uses an LLM-mediated approach where the orchestration layer reads API documentation, generates calls, and validates responses at runtime.

The Real Capability

• Pre-built certified connectors for finserv core systems (the beachhead reality).
• LLM-mediated connectors for long-tail APIs without pre-built integrations.
• Schema-drift detection — the model notices when an API response shape changes and flags before failure.
• Hallucination guardrails — every generated API call is validated against schema and dry-run before execution.
• Audit trail — every action ANi takes on a third-party system is logged, attributable, and reversible.

The Honest Failure Modes

• LLM-mediated calls are slower than pre-built connectors. ANi caches and optimises hot paths.
• Brittle behaviour on undocumented APIs is a real risk. ANi degrades gracefully and escalates to a human-in-the-loop.
• For regulated transactions (payments, trades), ANi requires a deterministic connector — never LLM-mediated.

This is what makes the integration story credible: we know the failure modes and we have engineering responses to each one.

Beachhead (Months 0–24)

• Mid-market financial services in EMEA — 250–2,500 employee firms.
• Sub-segments: independent asset managers, insurance brokers & MGAs, fintechs, regional & private banks.
• Why: regulated complexity creates integration pain; licence bloat is acute; procurement is burnt by lock-in; CFOs already buy AI in seat-priced packages.
• Geography: UK, Ireland, Benelux, DACH, Nordics — English-language, EU/UK regulatory regimes.

Expansion (Year 3)

• Mid-market financial services in North America & APAC.
• Same product, same methodology, different distribution & compliance footprint.

Vision (Year 4+)

• Adjacent regulated verticals: healthcare, legal, professional services.
• Long-horizon: any size, any industry, any location — the original ANi positioning, earned not assumed.

The acronym still stands — AI, No limits, industry & innovation — as the long-horizon brand promise. The wedge is what gets us there.

• Direct enterprise sales — small EMEA finserv-specialist sales team for strategic accounts and reference customers.
• Regional system integrators — delivery scale and local trust in each EMEA market.
• Finserv-specific events & thought leadership — Money 20/20 Europe, Insurtech Insights, regional CIO summits.
• Inbound digital — ROI calculator, sandbox demo environment, finserv-specific content.
• Reference customer / LOI program — target three paid pilots in 18 months as commercial validation.

• Named account team from sales through to operations — single point of accountability.
• Quarterly business reviews with each strategic account, including a documented licence-savings audit.
• Dedicated implementation pod for the first 90 days — de-risks first deployment.
• Self-service operator portal for day-to-day usage; community for peer support.
• Finserv user group & quarterly councils — customers shape the roadmap.

• Frontier AI providers: Anthropic, Google, Microsoft — token pass-through; multi-model resilience prevents single-vendor exposure.
• Cloud & infrastructure: AWS, Microsoft Azure, Teraco and regional EMEA colocation for data sovereignty.
• System integrators: regional and mid-market focused SI firms — delivery scale where ANi cannot reach directly.
• Finserv ecosystem: integration partnerships with core banking, claims, custody, KYC and AML software vendors.
• Finance partners: seed and pre-seed VCs in fintech and B2B SaaS to fund expansion phases.
• Universities & education institutions: talent pipelines and applied research collaborations.
• Phase 2 (post Series A): hardware manufacturers for the secure terminal device.

• R&D: orchestration platform, integration framework, governance & observability — software only.
• AI inference: frontier model API costs — token pass-through to customers, margin-protected.
• Infrastructure: cloud compute, EMEA data sovereignty, security & compliance tooling.
• People: engineering, finserv domain experts, sales, customer success, implementation.
• Go-to-market: EMEA finserv sales & events, channel enablement.
• G&A: compliance (GDPR, DORA, MiCA), legal, finance, IP protection.

Profile: software-only MVP keeps the burn predictable. Token costs flex with revenue (margin-protected pass-through). Hardware capex is explicitly deferred to a Series A+ decision, contingent on commercial validation in the software business.

• Seat subscription: the core mechanic. Predictable per-user / per-month across Essentials, Professional, Enterprise tiers, with generous included token allowances.
• Token overage: transparent, capped, alertable usage beyond included allowance. Never a surprise on the invoice.
• Outcome-based add-on: optional gain-share pricing tied to documented hours saved or licences retired. Reserved for sophisticated buyers.
• Implementation engagements: fixed-fee, scope-bounded, ANi-methodology delivery.
• Handover & exit-assistance package: paid offering that monetises the no-lock-in commitment and builds trust at the contract stage.
• Shared licence savings: gain-share arrangements where ANi captures a percentage of demonstrated software cost reduction it unlocks.

• Beachhead defined: ✓ — Mid-market EMEA finserv, 250–2,500 employees.
• Model-layer honesty: ✓ — Orchestration on top of frontier models, not a proprietary engine.
• Pricing a CFO can sign: ✓ — Seat-based with capped token overage and optional outcome pricing.
• Defensible moats: ✓ — Trust (no-lock-in), integration depth, methodology IP.
• Hyperscaler defence: ✓ — Mid-market focus, vendor-neutral, services-led.
• Two or three paid pilot LOIs: ☐ — Target for the next 18 months. This is the remaining commercial validation.

Five of six are addressed in this revision. The sixth — paid pilot LOIs — is the next-12-month execution priority and is what converts this from an investable thesis into a closed seed round.