KsefChase — KSeF cockpit for Polish biura rachunkowe

1. One-liner

Multi-client KSeF + bank-feed cockpit that chases missing invoices for Polish biura rachunkowe.

2. Trend signal — why now?

Poland’s mandatory e-invoicing system KSeF (Krajowy System e-Faktur) went live for large taxpayers on 1 February 2026 and extends to ~3M small businesses on 1 April 2026. From 1 August 2026, every B2B bank transfer must carry the KSeF identifier in the payment description, including all split-payment (MPP) transfers above PLN 15,000 for Annex-15 goods. Penalties of up to 100% of VAT kick in 1 January 2027.

The current state of the rollout is messy:

• Only ~13% of firms have implemented KSeF and 18% don’t intend to at all as of early 2026 — see PIT.pl reporting.
• 45% of surveyed businesses rate the rollout negatively, 36% express serious concerns (Bankier.pl).
• KSeF itself sends no notifications — recipients must poll. Three-plus thin SMS-alert micro-apps already exist (KSeF Alert, KSeF Powiadomienia, Alerty KSeF) at ~20 PLN/mo, proving willingness-to-pay for a pure notification layer.
• Polish accountant trade press (Infor.pl) reports księgowi explicitly refusing to take on responsibility for KSeF errors at clients, demanding tooling that keeps the line bright.
• JPK_V7(3) schema went live the same day, adding OFF / BFK / DI marker codes that have to match KSeF reality on every monthly VAT return.

Provenance:

• Signal 1 (demand): “KSeF paraliżuje biznes” — księgowa frustration coverage at Infor.pl ksiegowosc.infor.pl/ksef/7549129 — observed 2026-05-08; only 18% of small biz issue regularly per Bankier.pl 9110827, observed 2026-05-08.
• Signal 2 (feasibility): Official KSeF 2.0 production API live since 1 Feb 2026 (ksef.podatki.gov.pl) and PolishAPI PSD2 covers 31/281 PL banks via aggregators (openbankingtracker.com/country/poland) — observed 2026-05-08.
• Signal 3 (economic): SaldeoSMART made KSeF retrieval free Aug 2025, Comarch Optima/Symfonia compete on per-document KSeF tiers (50–600 PLN/mo); 70K biura rachunkowe + 3M JDG = a regulated spend bucket already in motion (we-tax.pl, ksef.pl) — observed 2026-05-08.
• Category: Regulatory arbitrage.

3. The opportunity

Phase 2 turns every księgowa in Poland into a multi-tenant KSeF operator overnight. The Polish accounting-software market reacted by adding “send & retrieve invoice to/from KSeF” features in client-side invoicing apps (Fakturownia, inFakt, wFirma, ifirma) and integrating document-level OCR + KSeF retrieval in biuro tools (SaldeoSMART, Symfonia, Comarch Optima). What none of them solved:

1. The chase loop. A biuro serving 100 JDG clients gets 5–15 missing-invoice incidents per week — vendor didn’t issue, client didn’t approve, bank transfer went out without a matching KSeF document. Today this is detected at month-end during JPK_V7 prep, when correcting it is 3× the work.
2. The Aug-1-2026 transfer rule. Every payment file must carry the KSeF ID. ERP monoliths will get there; the long tail of small biura, who export bank transfer batches as CSV/MT940 to PKO BP / mBank / Santander, will not — and they’re the ones with the deadline.
3. The JPK_V7(3) OFF/BFK/DI marker discipline — system has to read invoice metadata + KSeF events and pick the right code; mismatches trigger urzędowe wezwania.

A focused tool that lives between KSeF, the bank, and the client’s WhatsApp beats incumbent ERPs because the incumbents are still selling per-document KSeF packages while the actual unmet pain is reconciliation, chasing, and transfer-file compliance.

The wedge: “You will never lose another invoice between KSeF and your bank again — and we’ll auto-block transfers that don’t carry the right KSeF ID.”

4. Target market

• Primary customer: Owner-operator of a biuro rachunkowe in a Polish Tier-2/3 city (Lublin, Bydgoszcz, Rzeszów, Toruń, Gdynia, Olsztyn class), 1–3 księgowe staff, 30–150 active clients (mix of JDG + small sp. z o.o.), monthly billing 200–400 PLN per JDG / 500–1,200 PLN per spółka.
• Why they buy: “I cannot babysit 80 NIP-y in KSeF every morning. If I miss one missing invoice, I eat the JPK_V7 correction. From August I have to stop transfers without KSeF IDs or my client’s bank rejects them. I need a single tablica that screams when something’s off.”
• Rough TAM: 70,000 biura rachunkowe in Poland (we-tax.pl / ksef.pl). Realistic SAM: 25,000 biura serving 30+ clients, sub-Warsaw market. At PLN 199–599/mo per biuro, SAM is roughly PLN 75–175M ARR (€18–42M / $20–45M).
• Why now for them: Apr 1 2026 phase-2 mandate hit two months ago in this scenario, Aug 1 2026 transfer-ID rule is 12 weeks out, JPK_V7(3) penalties become real Jan 1 2027.

5. Product sketch (MVP)

• Multi-client KSeF inbox — single login pulls invoices for all clients the biuro is authorized on (mass-authorization model already supported by KSeF API).
• Bank-feed match (PolishAPI / Salt Edge / Tink) — auto-pair KSeF invoice → bank transaction → JPK_V7 line; flag pairs that don’t match.
• Missing-invoice radar — recurring vendors / contractual relationships not seen this month → WhatsApp/SMS nudge to the JDG owner or directly to the vendor with one-click “we’re still waiting on this invoice”.
• KSeF-ID transfer guard — generates Elixir-0/MT940 batch payment files for PKO BP, mBank, ING, Santander, Pekao with the KSeF ID embedded in the title, blocks export if any line is missing the ID (mandatory from 1 Aug 2026).
• JPK_V7(3) auto-marker — picks OFF / BFK / DI per Ministerstwo Finansów rules, ready to export each month.
• Anomaly digest for the księgowa — every morning at 7am, a one-screen tablica: “8 invoices below threshold, 3 vendors silent, 2 clients missing KSeF approval, 1 transfer blocked”.
• Client app (PWA / WhatsApp) — JDG owner gets a 30-second daily ping: “Approve these 3 incoming invoices in KSeF. Reply OK.” 90% of approvals happen in WhatsApp, never in software.
• Audit trail per NIP — exportable history of every approval, transfer, and KSeF event for tax-office disputes.

6. AI angle — what’s load-bearing

• Vendor-recurrence model — LLM + simple stats over 6+ months of invoices learns each client’s expected vendor list and cadence. This is the missing-invoice radar; rules-based heuristics flag too many false positives at biuro scale.
• OFF/BFK/DI classification — the KSeF “designation” rules from JPK_V7(3) are messy edge cases (offline mode, B2C, foreign vendor, paper). LLM with a forced-output schema gets 95%+ on this, hand-rules can’t.
• Polish-language WhatsApp client agent — auto-drafts “panie Tomku, w marcu nie wystawił pan faktury dla X — czy faktura się opóźnia?” in proper Polish + handles owner replies. Needs LLM, can’t do this with templates.
• Disputes & explanations to urząd skarbowy — auto-drafts the “wyjaśnienie” letter when a JPK_V7 mismatch arises, citing KSeF event log.

If you remove the AI: vendor-recurrence becomes manual rules, the WhatsApp chase is templates that the JDG owner ignores, and OFF/BFK/DI classification reverts to a checklist the księgowa fills monthly. The product collapses back into “KSeF Alert with extra steps.” AI is load-bearing.

7. Localization angle

This is Poland-first, Poland-only by design — that is the wedge. Specifically:

• Polish-language LLM prompts and Polish UI (księgowe over 40 don’t tolerate English-first products).
• PolishAPI PSD2 integration (NOT generic Plaid Europe — coverage of mBank, PKO BP, ING BSK, Santander BP, Pekao, Millennium is the unlock).
• Elixir-0 / Elixir-OnLine transfer batch file format — mandatory for the KSeF-ID-on-transfer enforcement; foreign competitors won’t ship this.
• JPK_V7(3) schema is Poland-unique; OFF/BFK/DI codes don’t exist outside Poland.
• WhatsApp + SMS routing — Polish JDG owners do not log into web apps; SMS still dominates outside Tier-1 cities.

Pricing must work at PLN 199–599/mo, which forces lean ops; this is fine because revenue is recurring and biura churn very low (1–2%/yr historically per Polish accounting industry data).

8. Business model — path to $1M–$5M ARR

• Pricing (per biuro rachunkowe, monthly):
  • Starter — 199 PLN/mo — up to 25 active client NIPs, 1 księgowa seat.
  • Growth — 399 PLN/mo — up to 75 client NIPs, 3 seats, WhatsApp chase included.
  • Pro — 699 PLN/mo — up to 200 client NIPs, 6 seats, KSeF-ID transfer guard for unlimited bank accounts.
  • Per extra NIP above tier: 6 PLN/mo.
• ACV: Blended ~5,000 PLN ($1,250 / €1,170) per biuro per year.
• Path to $1M ARR: ~800 paying biura at the Growth tier (≈1.1% of the 70K biuro market).
• Path to $5M ARR: ~3,500–4,000 biura (≈5% of market) plus a JDG self-serve tier at 39 PLN/mo for direct sales without a biuro (~10K JDG = $470K extra).
• Expansion path: (a) per-NIP overage as biura grow, (b) JDG self-serve adjacent product, (c) biuro-branded client portal as upsell, (d) tax-office-correspondence (e-Doręczenia) module once enforced 2027.
• Margins: API costs are the wildcard — KSeF API is free, PolishAPI is free at AISP licensee level (need partnership with Salt Edge / Kontomatik ~€0.30/connection/mo), LLM at ~₪0.20/client/mo. Gross margin 78–85% at scale.

9. Go-to-market wedge — first 100 customers

1. Direct outreach to “Złota Setka” biuro listings. Scrape Stowarzyszenie Księgowych w Polsce, Polski Związek Biur Rachunkowych, and clutch.co/pl/accounting (≈4,000 listed biura). Personalized 90-second Loom in Polish showing one of their clients’ KSeF feed live (using public NIP data) → “this is what your morning looks like in 8 weeks.” Aim 4–6% reply rate, 1.5% conversion = ~60 paying biura.
2. YouTube + LinkedIn co-marketing with księgowa influencers. ~12 księgowi influencers on YouTube/PL-LinkedIn (Aleksandra Kasińska, Anna Kuczyńska-Cesarz, “Pomysłowi Księgowi”, “Tax Care”) have 5K–80K subscribers and pump KSeF content. Sponsored deep-dive + affiliate code (15% rev-share for 12 months) = ~30 paying biura.
3. The “Save Aug 1” campaign. Buy Google ads on “elixir KSeF mt940”, “PKO BP KSeF przelew”, “split payment KSeF identyfikator” — niche commercial intent, low CPC (PLN 1–3). 4-week landing-page campaign with biuro-branded ROI calc → ~25 paying biura.
4. Conference push at Kongres Biur Rachunkowych (annual, ≈1,200 attendees) and SKwP regional chapters. Booth + 4 in-person demos × 8 cities = ~20 biura.
5. One-week-free audit hook. “Send us 1 client’s NIP, we’ll show you 30 days of missed invoices and pre-build your JPK_V7(3).” Converts to paid because they see actual gaps in their own books.

100 paying biura in 5–6 months is realistic given a 4-month product runway plus 2-month sales overlap.

10. Build complexity — justification

Medium. Off-the-shelf: KSeF API (open, documented), PolishAPI via Salt Edge/Kontomatik aggregator partnership, Twilio Polish SMS, WhatsApp Cloud API, OpenAI/Claude for Polish NLP. Custom work: Elixir-0 file generators per bank (≈8 banks × 1 week each), JPK_V7(3) export with new OFF/BFK/DI logic, multi-tenant authorization handshake with KSeF certificates, bank-feed-to-KSeF matching engine. A 2-person team (one Polish-fluent eng, one full-stack) ships v1 in 14–18 weeks; sales-led founder runs distribution in parallel.

11. Gating checklist

Gate	Pass?	Note
Legal in target market	✅	KSeF API public; PolishAPI PSD2 access via licensed AISP partner; no proprietary data scraping.
Ethical — no harm / dark patterns	✅	Helps SMBs comply; chase tone configurable; no adversarial collections.
Market exists (evidence above)	✅	70K biura, 3M JDG, mandate live; 3+ thin competitors already monetizing the notification slice.
1–5 person team can build this	✅	2 engineers + 1 sales-led founder. Polish-language operator strongly preferred.
Launchable with <$50K / ₹40L	✅	KSeF API free; AISP partner ≈ $1K/mo at small scale; LLM costs scale with clients; total seed budget $25–40K.

12. Feasibility score

Axis	Weight	Score	Notes
Problem intensity	20	18/20	Hair-on-fire for biura rachunkowe between Apr and Aug 2026; concrete penalty risk Jan 2027. The księgowa already loses a half-day per week to manual KSeF reconciliation.
Demand evidence	15	14/15	3+ paid notification competitors, 70K biura, 18% adoption gap = unmet, multiple Bankier/Infor articles documenting frustration. Discount 1pt because nobody yet pays for this exact multi-client power tool — assumption to test.
Build feasibility	15	11/15	Most pieces are off-the-shelf, but bank-file Elixir-0 generation and the multi-NIP KSeF certificate handshake have real depth; 14–18 weeks is a tight but real estimate.
Distribution clarity	15	12/15	Named lists (SKwP, PZBR), known księgowa influencers, conference circuit. Polish language barrier limits foreign founders, which is also why nobody’s swept this in yet.
Revenue mechanics	15	13/15	PLN 199–699/mo is squarely below biuro tools they already pay for (Comarch Optima can hit 600 PLN/mo just on KSeF tier). 800 customers to $1M ARR is achievable.
Time to first revenue	10	8/10	Annual prepay common for biuro tools; pilot pricing → 4–6 weeks to first paid contract.
Defensibility	10	4/10	Execution and bank-file integrations are the moat. SaldeoSMART or Symfonia could copy in 6–9 months once the wedge is proven; that’s the runway window to lock in 1,500+ biura.
Total	100	80/100

13. Qualitative modifiers

Founder-fit tags

domain-expertise-required — Polish-language, biuro rachunkowe workflow, KSeF/JPK_V7 regulatory detail. Cannot be done by a non-Polish team without a strong domestic co-founder. sales-heavy — first 100 biura is direct outreach + conference work, not inbound.

Key assumptions to validate (3–5)

1. Assumption: Biura rachunkowe will pay PLN 399/mo when they already pay Comarch/Symfonia/SaldeoSMART. How to test: 25 in-person/Zoom interviews with biura serving 30–150 clients. Ask: “If this saved you 6 hours/week and prevented one JPK_V7 correction per quarter, would you pay PLN 399/mo on top of your existing stack? Sign LOI.” Goal: 10 LOIs in 3 weeks.
2. Assumption: PolishAPI bank-feed coverage via Salt Edge/Kontomatik covers 80%+ of biuro client deposit-account base. How to test: Pull supported-bank list, cross-reference with NBP small-business banking-share data; check live with one biuro’s anonymized client list.
3. Assumption: Biuro is allowed (under PL law + KSeF terms) to act as the KSeF retrieval agent for clients en masse with mass authorization. How to test: Confirm with 3 doradcy podatkowi (tax advisors); read MF guidance on biuro-rachunkowe authorization model (already documented per wFirma + Symfonia + Fakturowo).
4. Assumption: WhatsApp/SMS chase to JDG owners actually moves the needle vs. status quo of księgowa phoning the client. How to test: Run a 4-week pilot in 2 biura, measure missed-invoice resolution time before/after.
5. Assumption: The Aug 1 2026 KSeF-ID-on-transfer rule is enforced by banks (not just Ministry of Finance posture). How to test: Check published technical specs from PKO BP / mBank / ING for August release; talk to 2 bank partner-program contacts.

Risk flags

1. Incumbent bundling risk: SaldeoSMART made KSeF retrieval free in Aug 2025; Comarch/Symfonia could fold a “chase + reconcile” feature into existing bundles. Mitigation: ship the Elixir-0 bank-file generator + Polish WhatsApp chase agent as first-mover differentiators that incumbents are slow to copy.
2. Regulatory churn: The KSeF rollout has already slipped twice (originally July 2024). Phase-2 dates (Apr 1 / Aug 1 2026) could move; that pushes urgency out, not away. Mitigation: price the urgency tier as “phase-2 ready package” so deferral helps retention.
3. Bank-API access risk: PolishAPI has uneven coverage; AISP licensing is non-trivial. Going via Salt Edge / Kontomatik is correct but adds margin pressure. Mitigation: launch with top-5 banks (84% market share) + manual MT940 import for the rest.
4. Founder language requirement: Non-Polish founder will fail at the 25-interview validation step. Mitigation: explicit go/no-go on Polish-fluent operator on the founding team.

14. Structured verdict

Score:                  80/100
Verdict:                STRONG GO
Confidence:             High
Best-fit builder:       Polish-fluent operator with biuro rachunkowe network or accounting-software past, paired with one full-stack engineer
Time to revenue:        4–6 weeks (paid pilots), 14–18 weeks (production v1)
Capital to launch:      $25–40K (PLN 100–160K) — mostly Salt Edge/Kontomatik, infra, LLM, conference booths
Top 3 assumptions to validate first:
  1. 10 paid LOIs from biura serving 30–150 clients within 3 weeks of pitch (validates pricing + wedge)
  2. PolishAPI/Salt Edge bank coverage exceeds 80% of small-biz deposit accounts (validates bank-feed match feasibility)
  3. WhatsApp/SMS chase to JDG owners reduces missed-invoice resolution time by ≥40% in 4-week pilot (validates AI angle)
Kill criteria:
  - Abandon if <3 LOIs from 25 cold biuro outreach calls in 3 weeks
  - Abandon if SaldeoSMART or Symfonia ships a multi-client chase + Elixir-0 KSeF-ID transfer guard before product v1 (i.e., before week 18)
  - Abandon if Aug 1 2026 KSeF-ID-on-transfer rule is officially deferred by ≥12 months

15. Next step — 1-week validation sprint

• Day 1: Build a 5-page Polish-language landing page (“KsefChase — w 8 tygodni nie zgubisz ani jednej faktury między KSeF a kontem”). Publish ROI calculator: “Ile godzin biuro traci miesięcznie na gonienie faktur?” Pull Złota-50 biuro list.
• Day 2: Cold-email 80 biura owners + warm-DM 20 księgowa influencers on PL-LinkedIn. Goal: 25 booked 30-min Zoom calls.
• Day 3–4: Run the 25 calls in Polish. Demo a clickable Figma of (a) multi-client KSeF inbox, (b) the missing-invoice radar tablica, (c) the KSeF-ID transfer-guard popup blocking a non-compliant transfer. Ask for PLN 199 prepaid pilot deposit, refundable in 30 days.
• Day 5: Decide go / no-go.
  • Go if ≥10 prepaid pilots collected at PLN 199.
  • No-go if ≤4 pilots — rework wedge or geography.

The pilot deposit is the falsifiable signal — biura are tight with money; PLN 199 in cash beats any “I’d buy this” handwave.