Portfolio

Jev

Here we use Jev to judge your decision making. I want to show you how it works, and how can we apply something like Jev to other projects.

Telemetry

When you submit a response, we send it to Jev and Llama 3.3 along with three separate grading criteria. Both AI models choose the criteria that fits your answer best, then the table below compares how they performed.

Jev Llama
Model – Llama 3.3 70B
Time to answer – –
Its verdict
Cost of this answer – –
Cost of 1,000,000 submissions like this – –
Tokens read – –
Tokens written – –

The verdict

Nailed it –

Scraping by –

Psyche unravelling –

Certainty

Confidence –

How sure Jev is about its pick. If one outcome scores far above the other two, confidence is high. If all three are close, it is low. 0% means a three way tie and 100% means no doubt at all.

Genuine answer? –

Before judging, Jev checks that you actually tried to answer. It asks itself a single yes/no question and scores the chance the answer is yes. That kind of question is called a "Noul". A real answer, even a silly one, scores high, and gibberish scores near zero. Below 35% the game asks you to try again.

Raw response 
Nothing yet.

Prices are published list prices (checked 29 September 2026). We limit Llama to a one word reply, which keeps the comparison fair.

A fun way to learn Jev

In this game you write how you'd respond to a fictional situation. Two AI models assess your answer, and the story continues after judging your result.

I built it to learn and share how Jev works. And what is special about Jev exactly? It's a type of AI model that ONLY scores text against criteria you provide, rather than writing a conversational response like Chat GPT or Gemini.

What does that mean?

For each scenario in this game, I wrote three sets of criteria outlining what I think makes a good, mid, or crazy answer. Jev looks at the scenario and your response against those three criteria, then returns how likely your response matches the outcomes. It also checks whether you're answer is nonsense in order to stop spammy submissions.

The story itself is written by me. Jev's job is only to assess your choice, not invent what happens next.

For comparison, a more typical model (Llama 3.3) assesses your answer too. Llama is a more typical model, like what you might use normally for summarising things or answering questions. Here, I've limited it to a one word verdict so it isn't spending time and money writing an explanation.

The table then lets you compare their decisions, response times and estimated costs. These scenarios involve judgement, so it isn't necessarily a definitive test of which model is better.. It's more like a way to see how two different types of tools handle the same task.

Where could this be useful?

The game itself is a fun way to learn the tool, but the automation underneath is actually useful to anyone:

  1. Read some text.
  2. Check it against criteria.
  3. Decide what happens next.

That could mean routing a support ticket, flagging a comment for review, or approving if a request meets a policy. Your criteria and scenario will be unique, but an AI judging a quick and cheap outcome can be applied to anything.

For a decision making AI like Jev, the question is.. do you need AI to write a proper text response, or simply to make an assessment? If you don't need something written, a "System One model" model is worth testing, particularly when the task happens hundreds or thousands of times a day.

If your team spends time reading, sorting or checking the same kinds of information on repeat, I'd love to hear about it, and share how I can save you time and money. Email [email protected] or use the contact form.