Fin: 16.5% fewer hallucinations than its previous own model

What they built

Apex 2.0, Fin's own customer-service answering model, post-trained on NVIDIA Nemotron 3 Super with Fin's custom post-training system; it is now live for every Fin customer.

The result

Versus Apex 1.0: hard resolution +8.6%, resolution +1.7%, hallucinations -16.5%; live for every Fin customer.

The number covers fin production metrics across its customer base, company-reported.

“We built both Apex 1.0 and Apex 2.0 by starting with open-weight base models, then using our custom post-training system,” Fin said.

The layer: The learning loop

Reviewed corrections are kept as test cases and later used to train models.

A log of every correction and who reviewed it
Eval sets that grow from reviewed corrections
New versions, then post-trained models, from that signal

Read the guide to the learning loop.

Playbooks that use this case

When to own a model, and how to get there

More cases on the learning loop

Shopify

96%

lower estimated serving cost for one agent

A daily continual-learning loop for Sidekick's GraphQL agent: an expert-calibrated LLM judge scores production conversations, a self-healing pipeline repairs failures, and the repaired trajectories are trained into a smaller model (SFT, then GRPO against the judge) that replaces the frontier model (source).

Bridgewater Associates

13.8x

lower cost per task, with higher accuracy than the best frontier model

A Qwen3-235B model fine-tuned on Tinker with investor-labelled data to do six daily information-filtering tasks (e.g. which news matters to a macro investor) (source).

Atos

1% → 99%

valid structured output after fine-tuning a small model

A proof of value on Red Hat OpenShift AI: GPT-4.1 generated synthetic training data, then small models (Phi-4 Mini 3.8B, Qwen 2.5 1.5B/7B) were fine-tuned for multilingual structured extraction from social media, deployable locally (source).