Product 02
The Assessment & Placement Engine
A six-agent pipeline that takes a candidate from first touch to a CEFR-aligned, defensible result a teacher would have spent half a day producing.
The team that delivers it
What it is
A six-agent pipeline that takes a candidate from first touch to a CEFR-aligned, defensible result. Listening, reading, writing, speaking, and a free-form interview. The agents grade, calibrate against rubrics, and produce the explanation a learner (or their employer) actually reads.
Built for
EFO, the gate every paying learner passes through. It replaces the half-day a teacher loses to grading and explaining a placement test, and gives a result that an external auditor can defend.
Who else could use it
Language schools running English or any L2 placement. Corporate training providers grading language for internal mobility. Recruitment firms screening language for hires. Institutions running CEFR-mapped entrance assessments. Anyone whose grading is currently a human bottleneck producing inconsistent results.
What they get
The engine configured to their language and rubric, with their branded result report. Hosted, white-labelled, or on their own infrastructure.
Maturity
Customer 1 (EFO) shipped and serving daily. Productisable for English immediately. Retargetable to other languages with rubric and prompt work; the architecture does not change.
Pricing posture
Per-assessment SaaS for low volume. Annual licence for high volume. Setup fee per language pair.
Lives at
build-log.ts: Mar 'The assessment + placement engine'. Brain context: business-context.md, EFO-Technical-Reference.md.
What we'll be writing about it
- origin
The placement test that ate half a day of every teacher's week
Before the engine, every assessment cost a teacher three to four hours. We measured. Here's what changed when six agents took over.
For: Language-school directors and corporate L&D heads
- mechanism
Six agents grading the same candidate — and how they agree
Listening, reading, writing, speaking, follow-up: the four CEFR bands, the calibration step, and the scoring auditor that catches drift between them.
For: Heads of assessment at training institutions
- transformation
Why the explanation matters more than the score
What candidates and their employers actually do with a CEFR result, and how the engine writes the explanation a stranger can defend.
For: Hiring managers and recruitment firms using language as a hiring filter