Assessment scienceItem exposure is the quiet failure of a small question bank
A bank that is too small stops measuring knowledge and starts measuring who has seen the questions. The arithmetic is unforgiving and easy to check.
2 min read
Assessment software
Assessment software, built and operated for you
We design, build and run the systems that author, deliver and score assessments — and we stay on for the exam windows, which is where most of the difficulty actually is.
- Assessments delivered
- TODOAssessments deliveredTODO: over what period, and counted how
- Platform availability
- TODOPlatform availabilityTODO: measured over which window, excluding what
- Institutions served
- TODOInstitutions servedTODO: live customers only, not pipeline
- Median time to launch
- TODOMedian time to launchTODO: from signed contract to first live sitting
What we build
Six things we are asked for most
Rarely one on its own. Each of these usually arrives attached to at least one of the others.
Assessment platforms
Item banking, authoring, delivery and scoring as one system rather than four that synchronise overnight.
TODOTODO: outcome this replaces
Exam integrity
Session controls, evidence capture and audit trails built to survive an appeal, not just to deter.
TODOTODO: outcome this replaces
Integrations
LTI, SSO and roster sync into the systems an institution already runs, so assessment is not another island.
TODOTODO: outcome this replaces
AI-assisted authoring
Draft generation and item review with a human approving every item that reaches a candidate.
TODOTODO: outcome this replaces
Analytics and psychometrics
Item statistics and cohort reporting that tell you whether a question measured anything.
TODOTODO: outcome this replaces
Operate and support
Running the platform through exam windows, including the fortnight either side when it matters most.
TODOTODO: outcome this replaces
- Countries with active users
- TODO
- Languages supported
- TODO
- Data residency regions
- TODO
- Support coverage
- TODOTODO: hours, and in which zones
- Team members
- TODO
Why it is different
Assessment is a specialism, not a vertical
Three things that follow from treating it that way.
Assessment, not generic software
Scoring rules, item lifecycles and integrity requirements are the hard parts, and they are the parts a general-purpose build gets wrong last and most expensively.
Built for the exam window
Load is not uniform. A platform that is fine in March and falls over in May has not been tested against the only week that counts.
Evidence by default
Every consequential action is recorded in a form a regulator or an appeals panel can read, because reconstructing it afterwards is not possible.
How we work
The commitments that shape everything below them.
You talk to the engineers
The people who answer technical questions are the ones who wrote the code. There is no account layer in between.
Accessibility as validity
If a candidate cannot operate the interface, the score measures the interface. We treat that as a measurement defect.
Data stays where it should
Residency and retention are design inputs, decided before the first table is created rather than retrofitted.
No lock-in by obscurity
Your items, results and audit records are exportable in documented formats. Leaving should be possible.
Scoping
Get a realistic estimate, not a range
Tell us what you need to deliver and to whom. We will come back with a shape, a rough duration and the parts we think are riskiest — including the ones that argue against starting.
- A written scope you can circulate internally
- The assumptions the estimate depends on, listed
- Where we think you should not build
Recent writing
- Assessment science
Security and complianceWhat remote proctoring actually proves
Proctoring produces evidence about a room, not a verdict about a person. Treating its output as a decision is where institutions get into trouble.
2 min read
EngineeringWhy an assessment marketing site belongs on a static host
Static hosting removes a class of outage from the pages candidates see first, and costs less than the runtime it replaces.
2 min read
Tell us what you are trying to deliver
A short description is enough to start. If it is not something we should build, we will say so.