Skip to content

Phase two, on the same install

What it takes to make AI work here is what it takes to train AI on how you work.

Every serious attempt at AI of your own runs into the same wall, and it is never the model.

Capture first Traces and corrections run from your first week, because capture is the one part of this that cannot be backfilled later.
Then a benchmark you own Cases mined from real runs and real corrections, scored against your outcomes rather than a public leaderboard.
Training is downstream Fine-tuning is one thing you can do with that material. The evaluation environment is worth having even if you never train anything.

The order is forced

Nothing new gets installed between one phase and the next.

The reason to care about the order is money. Treat phase two as a separate programme and you buy the capture, the governance and the environment twice, and the second purchase never gets approved.

Fig. 05 · one environment, three phases

The same five pieces

Nothing new gets installed to do the second job.

Read each row across. The piece on the left is running for you today, and the piece on the right is the same one doing the training job later. Everyone else in this market builds a copy of your business and learns in the copy, even when the copy runs in your own cloud.

Fig. 06 · today's job, and tomorrow's

How an attempt is run safely

An agent rehearses against your real systems, and the write it would have made is caught.

An evaluation run reads from your live systems, so the state is genuinely today's. Its writes are intercepted and recorded rather than committed, so what gets scored is the action the agent would actually have taken.

Recorded replay

mock_all

Replays a recorded run with canned responses.

Real reads

real_reads

Real tools against real data with writes refused outright.

Real reads, captured writes

live_reads_captured_writes

The same real execution, except each intended write persists to the result instead of happening.

Write refusal during a training run is a single enforced seam rather than a convention, including a floor that refuses any write on execution paths where the capture intercept does not exist.

What gets scored

The trajectory, including the write, rather than the paragraph at the end.

Evaluation tools score an output against an expected string, which tells you almost nothing about an agent that can reach a defensible answer through a sequence you would never have allowed.

Tool selection

Did it reach for the right tools for this task, or improvise with whatever was nearest.

0.25

Write actions

The captured write, compared against what should have been written. The score most people never think to build.

0.20

Parameter accuracy

Right tool, right arguments. An almost correct amount or the wrong account is a failure here rather than a rounding error.

0.20

Output quality

A model judging the answer, pinned to a continuous scale when the run is gating a promotion, because a binary judge at a decision boundary is noise.

0.20

Tool order

Sequence matters when one step authorises the next. Checking eligibility after issuing the credit is a different act.

0.15

Around the scores sit two things worth naming.

The two questions anyone with an ML background asks

Both are fair, and the honest answers are better than confident ones.

How many episodes will you actually have?

Far fewer than a lab, and we do not need lab volume, because the unit here is not a rollout.

What nobody doing real enterprise work will ever have is the volume for training from scratch, because every episode is a real business action with a real cost.

What supplies the reward?

Today it is human.

Tying a decision to what the business measured afterwards is the next layer, and we will say plainly that it is a causal inference problem rather than a plumbing one: delayed, confounded, and cleanly identifiable only where a business will tolerate a holdout.

Stated plainly

Three things this is not, so nobody discovers them in month four.

Not a simulated environment There is no synthetic twin of your business here and no generated task set. The environment is your systems, which is the strength and also the reason volume stays modest.
Not a claim that corrections are labels A correction today is reviewed, promoted, and reaches every agent as a rule. Capturing corrections as trajectory level training pairs is work we have not finished.
Not a finished training stack Capture and rehearsal are live. Evaluation suites are in build. Fine-tuning and distillation on your own infrastructure are roadmap, and they are the last thing to arrive rather than the first.
capture and rehearsal, live evaluation suites, in build training on your infrastructure, roadmap

Start with the phase that pays for itself.

The useful first conversation is about the work in front of your team this quarter.