You are a product manager reviewing a redesign experiment. Separate evidence about dispatchers from evidence about a different user group.
Laya / ProductResearch preview
Product Decision Assist
Review evidence against a product claim; explore other decisions as experiments.
Start with the fine-tuned task, or explore an experimental decision below. Experimental fine-tune: results can be wrong; confidence is not accuracy.
This model was fine-tuned to classify evidence about a product claim. These fictional examples were used in training. The other tabs explore tasks not trained here.
Fine-tuned task · Choice
Does the evidence support this product claim?
Read the note, run the model, then compare its answer with the training reference.
Synthetic evaluation: 50.0% accuracy across 64 cases. These training examples are not a new test.
Choose a training example
A measured redesign result
Loaded original: synthetic training example · Edits are new inputs
Where this could fit: reviewing research before a roadmap discussion. The output describes the evidence; the product manager still decides what to build.
Try a change
Change the dispatcher comparison so the revised flow takes longer.
Compare the model’s answers
Run the example to compare the possible answers.
Live model result · Confidence included
Exact input and response for this run. Editing or resetting clears the pair.
Source: data/semantic-v3/admitted/pm-targeted-train.json · row targeted-slot004-o0 · source case v2-cb766ef0899300d1. The loaded original state and question were seen by selected update 256. Edited input is no longer the original training example. Synthetic wording may say ‘actual’; those events are fictional.
Decision needed · Experimental taskChoice
Which research step should we take before committing the next sprint?
You have one research slot. Choose between learning about the problem, testing a prototype, resolving gaps, or preparing a pilot.
Experimental workflow · Fictional example · Not a task trained in this fine-tune
1 · Review the inputs
Read the proposal and evidence, then review the team policy below. Choice selects one of the four research actions.
Try a change
Try the same proposal with no observed target-user problem.
2 · Read the model result
Compare the model’s answers
Run the example to compare the possible answers.
Live model result · Confidence included
3 · Your next step
Review the suggested action against the evidence and team policy before allocating the research slot.
Run an example to see the input and output for that run. Editing or resetting clears the previous pair.
Decision needed · Experimental taskScore
Fund a prototype now, or gather stronger customer evidence first?
Use the strength of the available evidence to inform the funding discussion. This score does not make the funding decision.
Experimental workflow · Fictional example · Not a task trained in this fine-tune
1 · Review the inputs
Read the research summary. Score rates the evidence on the four levels below.
0No supporting customer evidence is supplied.
1Customer requests are supplied, without directly observed friction.
2Customer friction was directly observed, but an intervention outcome was not measured.
3A comparison with a control or previous flow measured improvement after an intervention.
Try a change
Compare with a proposal based only on customer requests.
2 · Read the model result
See where the evidence lands
Run the example to place it on the defined scale.
Live model result · Confidence included
3 · Your next step
Use the evidence level to identify what is missing; the product manager decides whether to fund a prototype or request more research.
Run an example to see the input and output for that run. Editing or resetting clears the previous pair.
Decision needed · Experimental taskNoul
Run this study as planned, or revise it before recruiting?
Before spending a research slot, confirm that the study tests the intended users, product change and outcome.
Experimental workflow · Fictional example · Not a task trained in this fine-tune
1 · Review the inputs
Statement to assess: “The proposed study directly tests our product hypothesis.” Read the hypothesis and study plan below.
Try a change
Compare with a study of different users and a different outcome.
2 · Read the model result
Explore one clear statement
Run the example to see the model’s true / false probabilities.
Live model result · Confidence included
3 · Your next step
Review the alignment result. Resolve any mismatch in users, change or measurement before committing the research slot.
Run an example to see the input and output for that run. Editing or resetting clears the previous pair.
Try your own decision
Choose the Laya output type first. The form will change to collect exactly what that type needs.
Choice: your possible actions → one selected action. Score: your ordered levels → a position on that scale. Noul: your statement → its true/false probability.
If the queue is full, the connection drops, or controls stay disabled, reload to start again. This clears your inputs and results; copy any notes you want to keep first.
The model classifies evidence relative to an explicit claim and audience. It does not recommend what to build. Selected v3 accuracy: 50.0% on the small synthetic study.
On the same 64-case synthetic test: uniform four-option guessing 25%; base Laya 28.1%; this fine-tune 50.0%. No product rule-parser comparison was tested. Older synthetic task results: Score 33.0%, Noul 47.5% accuracy. These older tasks do not validate the experimental demos.
These workflows explore interpreting written evidence. We have not established an advantage over rules; structured fields and fixed numerical policies are usually better handled directly.
V3 targeted / fixed seed 7601. This model did not meet every study progress safeguard. 64 synthetic cases in 16 groups; four option orders are correlated. No expert or real-world validation. Training is frozen.
The app does not intentionally save your notes, call another language model or take actions. Use non-sensitive examples. On Hugging Face, inputs are sent to the Space for inference; hosting infrastructure may maintain operational logs.
Base model: Convai Innovations Laya. Independent fine-tune; no upstream or TypeSafe endorsement. Model cards include the complete evaluation and provenance.