State Machine
Every process has a finite number of states. Every transition is logged.
Process Mining · RPA · LSO:ATOM
Recover the margin your systems can't see.
The AIP Audit (Algorithmic Impact Process Audit) opens a process we carry to the end: financing, implementation, operations. A report is not enough.
- Average savings per client / year
- 396 tys. zł
- EU grants secured 2024–2025
- From first instalment with SmartFlow
- 0 zł
/anomalies // four real-world examples
This is how margin leaks in a 50-person company.
Four situations we hear in every audit. Real people, real hours, real money. See how many of these four you recognise in your own firm.
-
Accountant re-enters invoices from PDF to ERP
Every purchase invoice arrives by email as a PDF. The accountant opens it, reads it, manually types 12 fields into the system, verifies, saves. Average 8 minutes. 30 invoices per day.
- Hours per month
- 88 h
- Cost per month
- 4400 zł
What QA10 does OCR + validation + auto-import. The accountant approves, not re-types.
-
Warehouse worker counts stock manually at end of day
Every day at 16:00 the warehouse worker walks the aisles with a clipboard, counts 220 SKUs, enters data into Excel, sends to the manager. 45 uninterrupted minutes. 22 working days.
- Hours per month
- 16 h
- Cost per month
- 1900 zł
What QA10 does WMS with barcode scanning + sync every 15 minutes. Everyone sees live stock levels.
-
Customer service answers the same 5 questions
30 emails per day, of which 24 are: where is my parcel, when is the invoice, how to return, password reset, availability. Each email 6 minutes. 4 hours a day go in circles.
- Hours per month
- 88 h
- Cost per month
- 7900 zł
What QA10 does AI assistant in the inbox — handles FAQ, routes the rest to a human. The human does what AI cannot.
-
Complaints resolved reactively, not preventively
A complaint arrives, 4 people gather for a meeting, analyse, call the supplier. 8 hours × 4 people = 32 hours per case. Average 3 complaints per month.
- Hours per month
- 96 h
- Cost per month
- 12 500 zł
What QA10 does Anomaly prediction on production data — batch rejected before dispatch, not after a complaint.
See it in action // real process simulation
This is what an order process looks like. Before automation and after.
Simulation of an order flow from email receipt to ERP record creation. Numbers from the Lider Przemysłu case — average cycle 27 minutes for 4 people, reduced to 2:30 for 0 people after deployment.
Email received
Outlook
PDF opened in email
Outlook + Excel
Manually re-entered into ERP
Symfonia · ręcznie
Verification + correction
Symfonia · weryfikacja
Invoice saved
WMS
Record in ERP created
Excel reconciliation
Email received
Inbox webhook
Document AI parse
QA10 pipeline
Rule validation
rules engine
Auto-create ERP
Symfonia REST API
Notify team
Slack #ops
Audit log + done
QCare telemetry
How we work // from diagnosis to ongoing operations
Four stages. Each has a definition of success. Each has a gate validation.
We don't sell a report or a one-off deployment. We guide you through the full cycle: from the first audit through funding and deployment to production operations. One team, one accountability.
-
Stage 01
AiP Audit
Process Mining on data from your ERP, CRM, and operational systems. In 1–3 mies. we show you where you are losing margin and exactly how much.
- Process map with unit costs
- Leak list with calculated cost of inaction
- Solution architecture — 3 variants (Core / Performance / Enterprise)
- Grant application dossier (if you qualify for Dig.IT/FENG)
-
Stage 02 (optional)
Funding
Two modules from Annex 4 of the AiP Agreement. Call Monitoring (9 months, 750 PLN one-time) — a certified EU Funds expert identifies active EU/national/Norwegian/Polish-Japanese calls for funding the deployment. Funding Application (fixed-price after positive qualification) — full dossier with feasibility, financial, and risk analysis.
- 9-month monitoring — alerts 90/30/7 days before call deadlines
- Funding application after positive expert qualification
- Full dossier (feasibility study, schedule, budget, declarations)
- Representation in the evaluation process — expert comment analysis, supplements, negotiations
-
Stage 03
QDeployment
Three variants: Core, Performance, Enterprise. Cover 3–5 processes selected from the audit. Scope depends on losses from Stage 1, funding from Stage 2, and management decision.
- Automated processes deployed to production
- Technical documentation + runbook + disaster recovery plan
- Competency transfer — 2 people at your site trained for maintenance
-
Stage 04
QCare
After go-live we maintain the solution in production; SLA with contractual penalty. Without this the deployment starts degrading in the third month and losses return.
- Uptime ≥ 99.5% annually
- Defect reduction each quarter (target: −10% errors q/q, Performance+)
- Quarterly ROI recalculation — you know what returned how much, in PLN
Results // five deployments, four public case studies
What we calculated together with our clients.
Five deployments, 4,944 hours freed in portfolio per month, PLN 13.9M in grants secured, average payback 19 months. Below are four engagements we closed from audit through QCare.
-
Industry: Serial manufacturing · 50 FTE · project in progress
Lider Przemysłu
PLN 712K in annual benefits hidden in repetitive processes- Projected payback
- ≈14 months
- Hours to be freed
- 2,592 h
AiP Audit Growth Platinum (completed) + QDeployment Performance (in progress). Designed architecture: anomaly detection on 4 SCADA lines — AWS Bedrock + Anthropic Claude. Root-cause hypothesis engine on Vertex AI + Gemini 3.1 Pro. Projected TDC from 71% to 89%, Go-Live planned Q3 2026.
Full case study -
Industry: GameDev · FERS 1.4 · MEN Area 6
Educational gaming studio
PLN 5.45M in funding with zero own contribution- ESF+ (EU, 82.5%)
- PLN 4.50M
- Own contribution
- PLN 0
Funding: identification of FERS 1.4 Area 6 — MEN programme with zero own contribution (EU 82.5% + state budget 17.5%). Full application dossier: 3 advanced strategy games teaching analytical thinking.
Full case study -
Industry: Retail · 75 stores
Retail chain — ESL
91% reduction in price-tag management time- FTEs recovered
- 4.2 FTE
- Rollout
- 6 weeks
ESL system vendor selection. Pilot of 5 stores, data validation, scaling. Integration with central ERP.
Full case study -
Industry: Professional services · audit + accounting
Audit firm
354% ROI in the first year- Onboarding reduction
- 78%
- Payback
- 4.3 months
Document AI for automatic classification of client source documents. KYC/AML workflow. Onboarding process standardisation.
Full case study
- 4,944 h/mo
- freed in portfolio (monthly)
- PLN 13.9M
- in grants secured
- 19 months
- average deployment payback
- PLN 396K
- average per client (annually)
/TESTIMONIALS // WHAT CLIENTS SAY
Client testimonials
"The audit uncovered 712,000 PLN in annual process losses that no one had counted before. QA10 didn't sell us software — it showed us our own problem in numbers."
"The deployment took 6 weeks. Quotation handling time dropped by 91%. We recovered 4 FTEs that now work on sales, not paperwork."
"354% return on investment in the first year. We had doubts about AI in a regulated industry — today I can't imagine going back to the previous operating model."
Contact // no sales rep · directly with the CEO
One team. One responsibility. Production results.
We tell you where you're likely losing margin, which grants you qualify for, and what a deployment would cost in your context. We respond within one business day with a proposed time.
Book a callThe call is free. The diagnosis is a mutual decision — whether you see value in our method and whether we see a process worth fixing.