2025 — Present
Nexio Fleet GmbH
Co-founder · Strategy, Operations & Product
Berlin & Brandenburg, Germany
German Mietwagen operators are audited by the LABO on driving and rest times. Most run that obligation on spreadsheets and memory, reconciling shift exports by hand across the platforms their drivers work on. Miss the reconstruction and you don't get a fine — you lose the concession. Nexio turns the obligation into a workflow: pull shift data automatically, flag the gaps that would fail an audit, and produce the pack an auditor asks for.
The compliance pipeline
Replaced a monthly spreadsheet reconstruction with a five-stage pipeline an operator finishes in one sitting.
I mapped the audit obligation backwards from what the LABO actually asks for, then split it into stages that a machine and a human each own. The first three stages run unattended; the operator only sees the shifts that need a decision. Defining that hand-off point — what the software is allowed to correct silently versus what a licence holder must sign off — was the core product decision of the whole platform.
Shift and break review
Operators report roughly 60–70% less administrative time per compliance cycle.
The screen only shows shifts the rule engine could not clear on its own. Severity is encoded in the row itself, so an operator scans for colour rather than reading every line — the difference between reviewing 214 shifts and reviewing the six that matter. Every adjustment is written back as an audit trail entry, because a correction an operator cannot justify later is worse than no correction at all.
- Exceptions-first view
- Immutable audit trail
- Two-tier severity
Platform data ingestion
Removed the manual download step entirely — including for the platform that offers no API.
Driver activity lives inside the ride-hailing platforms, and not all of them expose it programmatically. I specified an ingestion layer that treats both cases as the same problem: whatever the source, the output is one normalised document per driver-day. That single shape is what lets the rule engine, the dashboard and the report builder each read the same data without knowing where it came from.
- Python · FastAPI
- Headless browser automation
- Firestore · Google Cloud
Commercial architecture and positioning
Restructured the offer into a compliance base plus paid modules, and moved the pitch off penalties entirely.
Selling compliance on the threat of fines put the product in the same mental category as an accountant's invoice — a cost to minimise. Reframing it around licence risk moved it into the category of things an operator cannot run without. Pricing followed the same logic: everyone pays for the compliance base, and the modules that create margin rather than obligation are sold separately.
Operators respond to "pass your next audit", not to "avoid a €X penalty". The first is an outcome they already worry about; the second is a number they discount.
- 3 pricing modules
- Full financial model
Reference
Reference letter available
From Aleksander Boski, CEO and co-founder of Nexio Fleet GmbH. Attesting to product ownership, strategic contribution, and operational scope during the platform's early-stage build.
Open reference letter →On the figures
Every diagram on this page was drawn from scratch for the case study. Nothing here reproduces a production screen, a customer record, or a commercial term.
- No client names. Pilot and reference operators are described by segment and region only.
- No platform names. Integrations are identified by type — with or without a public API — which is the part that actually shaped the architecture.
- No real records. Driver identifiers, shift windows and break times in Fig. 2 are invented and internally consistent, not sampled from live data.
- No commercial detail. Prices, revenue projections and shareholding are left out.