For teaching and for research groups

The academic plan

The knowledge base is free and unlimited on every plan, including no plan at all. For a lot of teaching that is enough on its own.

Full pricing

Why this is priced where it is

Not as a discount, and it is worth being straight about it. The academic channel does not pay for itself and is not expected to. What it produces is students who learn to ask where a number came from before they use it, and who arrive in industry already holding the habit.

That is a slow return and an honest description of the motive.

The workshop

Ninety minutes, a real research question from the group, solved live. Not a demo — a question somebody in the room actually has, worked through in front of everyone, including the parts where the data runs out.

[FORMAT DETAILS, WHAT YOU NEED FROM THE HOST, HOW TO BOOK ONE — fill in after the first one has actually happened. Do not publish a workshop offer before running one.]

Teaching materials

The thing we most want to exist: a set of exercises a lecturer can download and use next week without contacting us.

One of the three is written and is below. The other two are planned:

• An evidence-grading exercise: take five property values from five sources, grade them, defend the grading
• A selection exercise producing a decision record rather than an answer

[STATUS — exercise 1 written 20 August 2026 and usable as it stands; exercises 2 and 3 not written. STRATEGIA §4.5 step 3. This is the mechanism that spreads between universities without your involvement.]

Exercise 1 — find the comparison that is not a comparison

Time: 40 minutes. Group size: any. Preparation: none. What a student needs: a browser and the ability to read a table.

Hand each pair of students one comparison page from the grade index and one instruction: find every row on this page where the two numbers are not measuring the same thing, and say what makes each one different.

There are four kinds of answer, and a well-chosen page contains two or three of them:

1. Different standard. Two specified minima from two documents. A worked example is 2205, where the eleven per cent gap is mostly not the standard.
2. Different product form. Cold rolled sheet against Quarto plate, which on some grades moves the figure further than the choice of alloy does.
3. Different kind of value. A minimum a standard sets against a typical value a mill measured and printed without promising it — the distinction that costs the most money.
4. Different test orientation. The one nobody finds unprompted; it has a worked example too.

Why it works as teaching. The student is not asked to learn a fact. They are asked to distrust a table, which is the habit the whole subject depends on and the one that is hardest to transfer by lecture. It also fails usefully: a group that finds only the alloy difference has demonstrated the exact reflex the exercise exists to interrupt.

For the marker. Every claim a student needs is on the page in front of them, and the site carries the audit of its own limits at what we could not source, which is a reasonable second reading for anyone who finishes early.

Licence. CC BY 4.0, like the framework. Use it, change it, teach it without telling us.

What we would want back

Stated plainly, because a programme that pretends to be a gift is a programme nobody trusts.

Feedback from people who have not used it before. Students find the confusing parts faster than anyone else, and they say so.

Permission to describe how it was used — anonymously if you prefer, named if you are willing.

Citations, where the work warrants it. The framework is CC BY 4.0 and citable; the tool is a tool and should be cited as one if it did work in a paper.

For research groups rather than teaching

The research agent is the module that fits a group best: literature orientation with sources you can follow, and an export you can attach to a group notebook.

Read what our data covers and does not first. Coverage is strong on inorganic crystalline materials and absent on anything formulated, which for some groups makes this immediately useful and for others makes it irrelevant. We would rather you found that out on this page than after a purchase order.

What this page does not offer

• A site licence. Not available.
• Institutional invoicing. [CONFIRM — likely not yet.]
• On-premise deployment. Not built.
• Course accreditation or certification. No.
• A substitute for Granta EduPack. That is what a generation learned material selection on, it has a curated dataset behind it, and it is not what we are.