Fred - The User Research Shepherd
The platform for user research, analysis, and recruitment.
Start new thread

Fred - AI-orchestrated UX research with behavioural tracking

by•
Fred now turns UX research into an AI-orchestrated workflow: plan studies, recruit and manage participants, run tests, analyze sessions, detect patterns, and build reports in one place. This launch adds full AI orchestration, real-time and replay-based eye tracking, gaze heatmaps, smarter analysis, and a broader research suite for teams that need faster evidence without losing methodological control.

Add a comment

Replies

Best
Maker
📌
Hey Product Hunt 👋 I’m Imre, founder of Fred. We originally built Fred to solve a problem I kept seeing in UX research: teams were collecting more and more evidence, but the workflow around that evidence was fragmented. Planning lived in one tool, recruitment somewhere else, testing in another platform, analysis in spreadsheets or docs, and reporting became a slow manual process. This second launch is a big step forward for us. Fred now includes full AI orchestration across the research workflow. The goal is not to replace researchers, but to remove the operational drag around research so teams can move faster while keeping control over method, interpretation, and decisions. What’s new in this launch: • AI-orchestrated UX research workflows • Real-time and replay-based eye tracking • Gaze data and heatmap-ready session analysis • Broader support for research methods • Faster pattern detection and reporting • A more complete workspace for research teams, product teams, and agencies We believe UX research should be easier to run, easier to analyze, and easier to turn into decisions. Fred is our attempt to make that happen without reducing research to shallow AI summaries. I’d love your feedback, especially on where AI should help researchers most and where it should stay out of the way.

Behavioral UX + AI orchestration is such an underrated combo. Curious how accurate the intent tracking gets over longer sessions.

Maker
Thanks! That’s exactly the area we’re most excited about. Short sessions already give useful behavioral signals, but longer sessions are where intent patterns become more interesting because the AI can compare actions, hesitation, navigation loops, gaze or attention signals, and task progression over time. Accuracy improves when the system has more context, but we’re also careful not to treat every signal as certainty. The goal is to surface likely intent, friction, and confidence levels so researchers can validate faster, not replace their judgment.

The eye tracking plus behavioral replay angle is what separates this from tools that just analyze transcripts. The real test is where the AI draws the line between surfacing a pattern and flagging it as friction. How do researchers override or challenge those interpretations when the AI gets it wrong?

Maker
Absolutely. That is exactly why Fred does not present AI interpretations as black-box conclusions. Every signal we surface, whether it comes from eye tracking, behavioral replay, transcript analysis, or interaction data, includes a confidence level and a link to the evidences. Researchers can immediately see whether an insight is strong enough to trust, weak enough to dismiss, or ambiguous enough to investigate further. The AI’s role is not to replace the researcher’s judgment. It is to accelerate pattern detection and make the reasoning behind each flagged friction point transparent. Researchers can review the underlying evidence, challenge the interpretation, and decide whether it should remain an insight, be downgraded, or be ignored entirely. That human override is core to the product, because in UX research, context still matters more than automation.