History

Description

The History panel records every conversation between an agent and its learners. Superadministrators can review transcripts, apply filters (date, sentiment, topic, category), and export data for deeper analysis—revealing learning trends, common pain points, and opportunities to improve both course content and agent behavior.

Target Audience

Superadministrator

Features

Complete Chat Log

View every session the agent has had, with user identifiers or Anonymous labels if the agent is set to anonymous.

Transcript Viewer

Click any session to open the full learner question and agent response thread.

Multi‑Facet Filters

Narrow results by Date Range, Sentiment, Topic, or Category to focus your review.

CSV Export

Download filtered or full histories for analysis in spreadsheets or BI tools.

Insight Generation

Identify frequent questions, learning gaps, and tone alignment to guide prompt tweaks or additional training materials.

How to Use (step by step)

Open the History Panel

Sign in to Agentic OS
Click the agent’s name dropdown
Select History

Review the History List

See every chat session plus user identifiers (name, username, email, or Anonymous)

Inspect an Individual Chat

Click a session in the left‑hand list
The full transcript appears, showing the learner’s questions and agent’s responses

Apply Filters

Date Range – limit to a semester, week, or custom period
Sentiment – isolate positive or negative reactions
Topic – focus on specific subjects
Category – filter by predefined learner categories

Export Conversations

Click Export
A CSV downloads containing the visible (filtered) data for off‑platform analysis

Pedagogical Use Cases

Identify Learning Gaps

Filter by negative sentiment or repeated questions to spot concepts that confuse learners.

Refine Course Materials

High‑frequency topics may indicate areas needing clearer explanations or additional resources.

Assess Agent Tone & Style

Review transcripts to ensure the agent’s language aligns with institutional guidelines and learner expectations.

Prompt & Dataset Optimization

Use findings to tweak the system prompt or add new datasets that address common queries.

Accreditation & Quality Assurance

Exported histories provide evidence of learner support and instructional effectiveness.

By accessing, filtering, and exporting chat histories, superadministrators gain actionable insights to continuously improve both the Agentic OS and the educational experience it supports.