Cross-border tax platform for companies operating across the UK and Poland. Corporation tax, VAT (MTD + JPK_V7/KSeF), transfer-pricing documentation, and R&D credits (RDEC/SME + ulga B+R) on one engine. Same rubric-reasoning layer as GrantAI, widened to handle tax codes against HMRC and KAS simultaneously.
Pillar 05Planned Q3–Q4 2026UK + Poland cross-borderShares engine with GrantAI
What it does
— 01
Computes UK and Polish CIT positions together. Surfaces double-tax-relief claims, UK-PL treaty positions, and permanent-establishment risk. Produces one cross-border position — not two disconnected filings that the firm then has to reconcile.
— 02
MTD-ready for the UK. JPK_V7 and mandatory KSeF for Poland (from 2026). Ingests invoices from the firm's accounting system, maps VAT codes per jurisdiction, produces the returns in the required filing format.
— 03
Local file + master file drafting for UK-PL intra-group transactions. OECD-aligned. KAS and HMRC rubrics handled in parallel. Benchmarking inputs are flagged for human review, never fabricated — the reasoning is auditable.
— 04
UK RDEC, SME scheme, Merged Scheme — and Polish ulga B+R. Same rubric-reasoning engine as GrantAI: classifies qualifying activity, drafts the technical narrative, computes the relief, produces the claim pack. Shared architecture is why this capability ships first.
— 05
Reads the position the firm is already in — UK Ltd + PL sp. z o.o., single entity with a branch, expat payroll, contractor networks — and maps tax obligations per jurisdiction. Not tax-planning advice. Structured analysis of the position as it stands, written to be defensible on audit.
— 06
HMRC has been scrutinising R&D claims hard since 2024, and KAS audit intensity is rising. Every filing TaxAI produces carries the inspectable reasoning trail tax authorities now expect. Not aggressive optimisation — defensible positions with the working shown.
Who it’s for
UK+PL companies with genuine cross-border operations
E-commerce with Polish operations and UK sales. Consulting firms with entities in both countries. Tech companies with PL engineering and UK HQ. Currently juggling a UK accountant who doesn't know Polish rules and a Polish accountant who doesn't know UK rules — paying for both, reconciling neither. TaxAI is the cross-border layer that holds the full position.
Accounting firms servicing cross-border clients
Bring cross-border work back in-house. Licence the engine, apply your own judgment to the final review. Add a product line without building dual-jurisdiction expertise from scratch — and keep the client relationship.
How it’s built
Not a single prompt or a chatbot. A composed pipeline of specialised agents, each with a defined role, a documented toolset, and a declared failure mode.
agent/ingest-financialsPulls payroll, invoicing, and ledger data from both UK and Polish accounting systems (Xero, QuickBooks, Symfonia, Comarch ERP Optima). Maps to cross-border categories.
agent/classify-cross-border-positionIdentifies jurisdictional obligations, UK-PL treaty positions, permanent-establishment risk, and transfer-pricing triggers. Flags borderline positions for human decision.
agent/compute-taxRuns CIT, VAT, and R&D-relief computations per jurisdiction. Cross-checks treaty positions so double-counting is caught before it reaches a filing.
agent/draft-filingsProduces HMRC-formatted and KAS-formatted filings along with the supporting narratives. JPK_V7 and KSeF output compliant with the mandatory 2026 Polish format.
agent/qa-audit-readinessDefensibility gate. Audits every filing against current HMRC scrutiny standards and KAS audit patterns. Flags vague positions, unsupported reasoning, or gaps in the evidence chain.
Status
Status
Planned Q3–Q4 2026. The R&D-credit capability ships first — architecture is inherited from GrantAI, POC work carries across. VAT, CIT, and transfer-pricing modules follow.
Starting markets
UK + Poland cross-border, primary. UK-only and PL-only supported as secondary.
Compliance
HMRC rubrics (BEIS guidance, MTD specs, UK transfer-pricing local file format). KAS rubrics (ulga B+R, JPK_V7, mandatory KSeF format from 2026, PL transfer-pricing local file). Full reasoning trail for audit defence.
Code reuse
Significant architectural reuse from GrantAI. Same ingestion, rubric-parsing, drafting, and QA layers — tax codes replace grant-call rubrics.
Commercial model
Per-firm licence. Tiered by complexity — single jurisdiction, cross-border, group.
Register interest
enterprise@blackflake.com — be on the Q3 pilot list.
How we engage
Every deployment is a bespoke engagement. We don't sell seats or generic SaaS. We build agents against your workflows, your matter taxonomy, your regulatory context — then we operate the system with you under a monthly retainer. Your data, configurations, and outputs are yours and portable. The Blackflake engine is developed by us and runs under your retainer.
— 01
Scoped call to map your workflow, constraints, and success metrics. No charge. You leave with an architecture note and engagement estimate.
— 02
Fixed-price deployment on one narrow workflow. Real output on real work in weeks. Go / no-go before the full build.
— 03
Agent suite, workflows, sources of truth, QA gates, operator surface. Deployed in your tenant. Auditable.
— 04
Monthly retainer: new regulations ingested, agents tuned on live work, new workflows added. We develop the engine; your data, configurations, and outputs stay yours — always portable.