DoneThat

What is DoneThat?

DoneThat is an automated Activity Reconstruction and Productivity Analytics platform designed to capture and categorize professional workflows. The solution utilizes a cross-platform desktop agent to record screen activity and employ machine learning for the generation of structured work logs, Timesheets, and contextual data for Large Language Models (LLMs).

Key Capabilities

Visual Activity Capture: A desktop application (Windows, macOS, Linux) that performs periodic screen capture to reconstruct user activity. The vendor states the client application is Open Source to facilitate security auditing.
Automated Timesheet Generation: Converts captured visual data into categorized work blocks for billing and project management, intended to eliminate manual time entry.
AI Context & Memory: Integrates with external LLMs via API, MCP (Model Context Protocol), and automation platforms like Zapier. This provides a historical "contextual memory" of work history for use in AI-assisted tasks.
"Don" Productivity Coach: An interactive assistant that analyzes real-time activity to provide situational coaching, goal calibration, and interventions during periods of low focus.
Privacy Controls: Employs a "privacy-first" architecture allowing users to define recording parameters and select specific LLMs for data processing.
Workforce Visibility: Provides shared feeds and automated updates for remote teams to reduce the requirement for status meetings and build organizational trust.

Audience & Use Cases

Audience: Freelancers, software developers, and remote knowledge workers.
Use Case: Automating SKU-level or project-based billing without manual timers or activity logs.
Use Case: Identifying automation opportunities by analyzing the time-intensity of recurring digital processes.
Use Case: Providing long-term historical context to AI agents for personalized project assistance.

Technical Specifications