Headers

Systems that are in daily use

For each one we show what changed, how it was checked and where the limits are.

What we have built

One system for a client and two products of our own.

Chemical distributor

A day of retyping replaced by a check

Figures from supplier PDFs were retyped into spreadsheets by hand, taking one person a full working day. The software now pulls them out, flags what it is unsure about, and a person confirms the result before it is used.

  • Chemical distribution
  • Document processing
  • Language model
  • Data in the EU
Read the case study
document type detected correctly in 860 documents
100 %
mandatory fields extracted correctly
over 97 %
p95 per document
~40 s

Penmate

Faster marking. More context on how each paper was written.

Penmate assesses English writing against Cambridge English criteria and sends feedback to the student after submission. The teacher also gets more context on how the paper was produced: highlighted sentences, writing behaviour, and a short summary rather than a single verdict. The Headers team develops and runs Penmate.

  • English teaching
  • Cambridge English
  • Writing integrity
  • Assessment by a language model
Read the case study
to assess one piece
1.97 s
system capacity
thousands of pieces daily
guard reliability
4000+ tests

Headers

An agent that sits in the company Slack and remembers what was agreed

We built our own agent for the daily running of the company. It sits in our Slack, keeps a long-term memory, watches reminders and answers from company documents. It has been running with us since the summer of 2026 and we are turning it into a product.

  • Agent system
  • Slack
  • Long-term memory
  • Our own product
Read the case study
in use in our Slack
every working day
a reminder is delivered
within a minute
written to long-term memory
only with approval

Have a similar problem?

Tell us about it. We'll tell you whether and how AI can handle it.

Talk about your problem