Show HN: Meridian(PH #1) – Automatic AI Workjournal for Devs
Meridian is an open-source productivity tool that automatically captures screen activity to reconstruct your workday as a timeline, eliminating manual time tracking It generates AI-powered daily summaries, standup reports, and Jira worklog updates from actual on-screen activity rather than user input Built with Rust and Tauri, running as a lightweight background daemon (5-10% CPU, 0.5-3GB RAM) on macOS (Apple Silicon) and Windows 10/11 Privacy-first architecture: all activity data stays in a loc
Analysis
TL;DR
- Meridian is an open-source productivity tool that automatically captures screen activity to reconstruct your workday as a timeline, eliminating manual time tracking
- It generates AI-powered daily summaries, standup reports, and Jira worklog updates from actual on-screen activity rather than user input
- Built with Rust and Tauri, running as a lightweight background daemon (5-10% CPU, 0.5-3GB RAM) on macOS (Apple Silicon) and Windows 10/11
- Privacy-first architecture: all activity data stays in a local encrypted database; only anonymized summary data is sent to connected AI providers
- Supports retrospective queries ("What did I do three months ago?") and estimation analysis by logging actual work patterns over time
Why It Matters
Meridian represents a growing class of AI-powered productivity tools that shift the burden of documentation from the user to automation, which could significantly reduce administrative overhead for knowledge workers. Its local-first, encrypted approach to screen capture addresses a key concern in enterprise AI adoption—data privacy—making it a model for how activity-tracking tools can balance utility with security. The integration with Jira and standup workflows targets a high-friction pain point that affects millions of software developers and project managers globally.
Technical Details
- Stack: Rust backend with Tauri for the desktop tray application; Node 20+ and bun for frontend dependencies; pinned to Rust 1.93.1 via rust-toolchain.toml
- Architecture: Two-process design—a Rust daemon handles background capture and storage, while the Tauri tray manages UI and in-process screen capture, with automatic rebuilds on file changes during development
- Data pipeline: Screen activity is captured continuously, stored in an encrypted local database (~20 GB/month), with automatic pruning of captures older than 30 days; AI analysis runs through user-connected providers with only summary-level data transmitted externally
- Storage & performance: Approximately 20 GB monthly storage with automatic rotation; diagnostics are opt-out and stripped of identifying information before leaving the device
- Deployment: MIT-licensed, available via
git clonewith automated dependency installation (install-dev.sh) and development mode (dev-start.shlaunches both daemon and tray in watch mode)
Industry Insight
- The rise of local-first AI tools like Meridian signals a market shift toward privacy-preserving productivity automation, suggesting that enterprise adoption will favor solutions that keep sensitive data on-device rather than in cloud pipelines
- Tools that automate standup and worklog generation address a measurable productivity drain; organizations should evaluate whether deploying such tools at scale could recover hours of administrative time per developer per week
- The encrypted local database model combined with optional AI provider integration offers a replicable architecture pattern for other productivity tools seeking to balance automation capabilities with data sovereignty concerns.
Disclaimer: The above content is generated by AI and is for reference only.