The AI Pack
Describes version 4.37.1
Local AI analysis for your studies: session summaries, a cited findings report, grounded question-answering, and a study designer, all running entirely on your machine. Nothing about your study leaves your network.
The AI features need the ai entitlement on your license and the AI Pack installed (a separate installer that delivers the local model). With both in place the AI buttons appear; with either missing they simply don't.
To check what you have, open the … menu at the top right and choose About Ovo Logger. It names the app version and the AI Pack version separately, because they ship as separate installers on their own version lines, along with the model file in use. Both also appear in an exported diagnostics bundle.
AI output is labeled a draft everywhere it appears. Ovo Logger owns every number and timestamp - the AI writes only narrative. Every claim carries a clickable citation, and anything that can't be verified against the study's own data is dropped rather than hedged.
Installing the AI Pack
Where to get it. The AI Pack is a separate download from Ovo Studios, not part of the Ovo Logger installer and not something the app fetches for itself: the model is around 2.4 GB and this product does not reach the internet on its own. If your license includes the ai entitlement and you don't have the installer, email sales@ovostudios.com and we will send you the link.
Run the AI Pack installer on each machine that will do AI analysis. Like the app itself it is per-user with no administrator rights, and it places the local model next to the app. Updating the app does not remove the pack; the pack updates independently with its own installer.
Once it is in place the AI buttons appear by themselves. There is nothing to point at the model and no setting to turn on: the app looks for it beside its own program file, which is where the pack puts it.
The findings report
Open Analysis → AI findings (or ✨ AI Report in the review workspace) and click Generate. Ovo Logger summarizes each session locally, then drafts a plain-language findings report for stakeholders:
- What we tested · How it went overall · What got in people's way · Task by task · What we recommend.
- Every claim carries a citation like
[S2 @ 6:35 "…"]: click it to jump to that exact moment in the recording, even in another session. - A collapsed Researcher appendix carries the exact methodology detail - themes, severity, per-task scores - computed by the app, never paraphrased by the model.
Ask the study a question
💬 Ask takes a specific question: "Where did people struggle with checkout?", and answers it from only your study's data, with clickable citations. If the data doesn't answer the question, it says so rather than guessing. Ask is read-only: it never changes the study.
Study Designer
Describe your study in plain language - the product, what you want to learn, who your users are, and the Study Designer (on the Study Setup page) drafts a plan: tasks with success criteria, a screener, Quicklog labels, and a moderator guide.
- Fill in the brief and let it draft.
- Review and edit everything it produced - tasks, tags, screener, guide.
- Click Apply to project. Nothing is written until you apply.
AI history
Reports, answers and session summaries are kept. AI history lists everything the AI has produced for the study, so you can reopen an earlier report, compare it with a newer one, or find last week's answer. Regenerating never overwrites an old report - a finished report is a record of what you knew at the time.