Product

Two agents, one journey map.

The Generator creates the journey map from your research. The Optimizer reads it, finds what to change in the live version and, once a change is applied, adds the tracking events to your code.

Input

Research insights

Interviews, surveys, support logs, analytics.

Agent 01

Generator

Turns your research into a journey map.

Shared by both agents

Journey map

Steps, the AI’s role, handoffs, trust mechanisms and signals.

Agent 02

Optimizer

Learns from your users and proposes what to change.

Reads and updates it
Agent 01: the Generator

Turns research into a journey.

It goes from raw research to a journey ready to launch, in four steps:

InputInterviews, surveys, support logs, analytics.

  1. 01

    Ingest

    Reads interviews, surveys, support logs and analytics exports.

  2. 02

    Synthesize

    Clusters needs, pains and moments that matter, each linked to its source quotes.

  3. 03

    Design

    Proposes the journey steps, the AI’s role at each one, handoff rules and trust mechanisms.

  4. 04

    Instrument

    Defines the signals and the business metric to track at each step, plus the evaluation cases for launch.

OutputA journey map ready to launch: steps, the AI’s role, handoffs, trust mechanisms and the signals to track.

Agent 02: the Optimizer

Keeps the live journey improving.

It learns from your users, proposes what to change, and tracks each change once it’s applied:

InputSignals from the live journey: ratings, corrections, rephrased questions, abandoned sessions, unplanned handoffs.

  1. 01

    Collect

    Gathers ratings and corrections, plus implicit signals like rephrased questions, abandoned sessions and unplanned handoffs, and places each one on the step it belongs to.

  2. 02

    Decide

    Finds which step fails and which part of the system causes it, then chooses what to fix first, based on each step’s impact on the metrics that matter to your business.

  3. 03

    Improve

    Prepares the change, from prompts and examples to retrieval settings, tools, workflow or handoff rules, and tests it on real cases from that step. Only changes that do better are proposed.

  4. 04

    Apply and track

    Explains what changes for users. Once you approve, the change is applied to your product and the events to track it are added to your code, and the loop starts again.

OutputA tested change, applied after your approval, with the events to track it in your code.

How you work with them

Chat

Ask either agent anything in plain language, review the generated journey and approve changes in the conversation.

Dashboard

Your journey map with the health of every step, the status of every change, and its full history.

Shared foundation
  • Research insights
  • Journey map
  • Traces
  • Evaluation sets
  • Versioned configuration
  • Change history
Contact

Start with the journey your customers struggle with most.

Tell us about your platform and your users. In the demo, you’ll see how OttimAI maps your journey and measures every change.

Request a demo