24,000 files of incoming mail, sorted in minutes instead of weeks.
A regional housing company was drowning in letters and scans. Today Piko reads every item that comes in, recognises what it's about and files it against the right case — entirely in its own data centre.
- Industry
- Housing sector
- Portfolio
- ~ 24,000 units
- Deployment
- On-premise
- Module
- Piko Document Intelligence
- 0 h
- processing time / year saved
- 0 %
- of incoming mail classified automatically
- < 0 days
- from data export to first results
Mail piled up, deadlines slipped.
Three caseworkers spent whole mornings opening incoming mail, reading it, assigning it to the right property and transferring it into the core system by hand. Across ~24,000 units, an endless stream.
Terminations and objections were noticed too late — with real legal consequences.
Knowledge sat with individuals; holidays and sick leave brought cases to a standstill.
A cloud solution was never an option for data-protection reasons.
The core system was fixed — nothing was to be replaced.
Three steps, one continuous run.
No new portal, no relearning. The inbox stays the inbox — Piko sits in front of it and makes it readable.
- 01Ingest
Every inbound item in one place
Letters, PDFs, email attachments and scans flow into a single inbox. Piko takes the format as it comes — no pre-sorting, no renaming.
- 02Understand
Classification on your taxonomy
Reminder, termination, damage report, service-charge objection — every document is mapped to your own categories and the relevant fields are extracted.
- 03Route
Straight into the right case
The structured result lands at the matching property, tenancy and caseworker — with deadline, amount and context. No manual assignment.
We changed nothing about our system — and yet our mornings are back. The mail is already sorted before anyone even walks into the office.
Tailored to the portfolio — not off the shelf.
Four building blocks that together turn the inbox into structured, deadline-safe cases.
Custom document taxonomy
27 document classes, trained on the real portfolio — not on a generic standard.
Fully on-premise
Tenant data never leaves the data centre. No cloud, no third party in the data path.
Deadline detection
Dates and payment deadlines are read out and passed on as structured fields — nothing gets lost in the filing.
Integration with the core system
Results flow back via API straight into the existing housing-management software.
Sounds like your inbox? Let's talk.
Send us a day's worth of incoming mail — anonymised. Within a few days we'll show you what Piko recognises on your real documents.