Think Build Implement Repeat
London, UK +44 7367 067226
WhatsApp FOLLOW f in X
  1. Home
  2. Blog
  3. Evaluating AI Features Before and After Release
AI & Machine Learning

Evaluating AI Features Before and After Release

Without evaluation you are changing prompts and models on vibes. A practical evaluation setup for a product team, and what to watch in production.

Updated 2 min readBy SpiderHunts Technologies

Free estimateNo obligation

Get a free estimate

Tell us what you need. A senior engineer reads every enquiry.

Takes under a minute. We never share your details.

  • Free consultation
  • No commitment
  • NDA on request

Prefer to talk? Book a free 30-minute call →

Quick answer — TL;DR

Build a small labelled test set from real cases, run it on every change, and pair it with production signals such as correction rate and abandonment. Without both you cannot tell whether a change helped.

The short answer

Changing a prompt or a model without an evaluation set means you cannot tell improvement from regression. The change feels better because you wrote it.

A modest test set of real cases, run automatically, plus a few production signals, is enough to make changes accountable.

Build the test set from real cases

  1. Collect actual inputs from production, including the ones that went wrong.
  2. Have someone who knows the domain state the correct output.
  3. Keep the awkward cases, because they are the point.
  4. Run the whole set on every prompt or model change.
  5. Add each new production failure to the set.

Point five is what makes the set improve over time. Every bug becomes a permanent regression test.

Production signals worth watching

SignalWhat it suggests
Correction rateHow often output is wrong enough to fix
AbandonmentUsers giving up mid-task
Retry rateOutput not good enough first time
Escalation to a humanFeature not covering the case
LatencyAffects use more than teams expect

Correction rate is the most honest single measure for anything that produces content a user then acts on.

Do not evaluate only the average

An overall score can improve while the feature gets worse for a particular segment. Break results down by case type, customer size or language, whichever matters for your product.

Regressions concentrated in one segment are easy to miss in an aggregate number and are exactly what customers notice.

Keep it cheap enough to run

  • Small enough to run on every change without hesitation
  • Automated, with a clear pass or fail
  • Versioned alongside the code
  • Owned by someone specific
  • Reviewed periodically for cases that no longer matter

An evaluation suite that takes an hour and costs real money will be skipped under deadline, which is precisely when it was needed.

FAQ

Frequently asked questions

The questions readers ask us after this guide.

Still have a question?

Ask us directly — a senior engineer will get back to you.

Ask about your project

How large should an evaluation set be?

Small enough to run on every change without hesitation. Coverage of awkward cases matters more than volume.

What is the best production signal?

Correction rate, for anything producing output a user then acts on. It is the most honest single measure.

Why break results down by segment?

An aggregate score can improve while a particular segment gets worse, and that is what customers notice.

What should happen after a production failure?

It goes into the evaluation set permanently, so the same failure cannot return unnoticed.

Keep reading

More on AI & Machine Learning

Start here

Building an AI product and want a second opinion on the plan?

Tell us what the product does, who pays for it and where you are now. We will come back with an honest read on the architecture, the costs that scale badly, and what we would build first. No pitch for a rebuild you do not need.

  1. You tell us what you needTwo minutes on the form, or a message on WhatsApp.
  2. A senior engineer reviews itAnd comes back with questions, a realistic range and an honest view on fit.
  3. Free 30-minute scoping callWe talk through scope, options and a realistic estimate — with no obligation.
Free estimateNo obligation

Talk to someone who builds this

Send a short brief and we will come back with an honest view and a realistic range.

Takes under a minute. We never share your details.

  • Free consultation
  • No commitment
  • NDA on request

Prefer to talk? Book a free 30-minute call →