Parallel Hours is a PSP-inspired time and KPI platform for software developers. Log human and AI work side-by-side, track your velocity, and finally know how much of your output the AI is actually driving.
Time tracking is the foundation. KPIs are the point.
One click to track time against any task. Multiple concurrent timers so you can track AI and human work in parallel.
Velocity, estimate accuracy, defect rate, and AI leverage ratio — the metrics that actually tell you how productive you are.
Log prompt counts and AI-autonomous minutes per session. See your AI leverage ratio grow as you get better at delegating to the machine.
Connect Claude Code or any MCP-compatible agent. Start timers, log tasks, and increment prompt counts automatically from your AI tool.
Jira-style task IDs, story points, backlog-to-done workflow. Connects to GitHub Issues and Jira so you can track time against real tickets.
Filter by project, date range, and task type. Export to CSV for billing, performance reviews, or your own analysis.
Most developers have a gut feeling about their AI productivity. Parallel Hours measures it. The AI leverage ratio compares your AI-autonomous minutes to your total wall-clock time — so you can see whether you're getting a 2x multiplier or a 10x one.
Parallel Hours itself was built almost entirely with Claude Code. Every hour of development is tracked here, including the AI sessions. The AI leverage ratio on this codebase is over 8x — meaning for every hour of elapsed time, eight hours of work got done.
That's the story you want to tell your team, your clients, or your next employer. Track it.
Sample KPI dashboard
AI Leverage Ratio
8.2×
AI min / wall-clock min
Velocity
12 pts/wk
story points delivered
Estimate Accuracy
84%
actual vs estimated
T_wall this week
18 hrs
elapsed calendar time
AI Concurrency Ratio
Parallel Hours was designed and built by a senior software engineer and lead architect with a career spanning PSP, Agile, and Scrum delivery across large engineering organizations. That experience includes:
This entire platform — from data model to MCP integration — was built using AI-assisted workflows tracked in the tool itself, as a deliberate demonstration of what's now possible at that leverage ratio.
This product, by the numbers
Tech I work with
Also in progress
Alpha · May 2026DataSeal
A zero-trust data store with a separate, tamper-evident audit log — designed for environments where provenance and chain-of-custody matter. Working towards 21 CFR Part 11 compliance. Alpha release targeted May 2026.
Start free. Upgrade when you need MCP and integrations.
$0
forever
$18 / mo
or $180/yr — save 2 months
Free to start. No credit card required. Upgrade when you're ready for MCP and integrations.
Create your free account