OpenAI — Engineering Performance

Effective engineers behind OpenAI, against its pre-AI baseline. Each subject has its own: OpenAI's is 0.75 ETV / dev / mo, its first reading in Q1 2025. Per-engineer ETV divided by that gives a capacity multiple, and that multiple applied to the engineers active in the trailing 90 days turns it into engineer-equivalents. The line is the real headcount, so the gap between line and area is what the leverage is worth. Because each baseline is its own, every subject opens at 1.0x on its first day: multiples measure improvement and are not comparable between subjects.

Per-engineer ETV for OpenAI against the pooled 500 OSS Performance Index. Both lines are 90-day rolling averages scaled to a 30-day month, so they share one axis and can be read against each other at any point. Pick a window to zoom the chart to it. Latest reading: OpenAI is 278% above the index (14.55 vs 3.85 ETV/dev/mo). At the start of the tracked period the gap was 18% below.

Written summary of the work completed each month.

Where each repository stands: average performance per engineer per month over the last 90 days, with the rolling 90-day curve behind it. The range picks the window (past 90d): it sets how much of the curve you see, the Δ across it, and the span the work mix is measured over. Ranked highest first.

Each month's output split by type of work: Features (new value), Maintenance (sustaining systems), Tests, Docs, and Fixes (rework). The yellow line is output per engineer, so when it rises each engineer is delivering more, whatever the team size did. Unit: Engineering Throughput Value (ETV).

Each month's work split two ways. CapEx (capitalizable investment) is the work that builds the asset: Features, plus the Tests that prove it works and the Docs that explain it. OpEx (operating expense) is keeping it running: Maintenance and Fixes. The yellow line is the CapEx share, so a rising line means more of the month went into building new rather than sustaining what exists. Unit: Engineering Throughput Value (ETV).

Who shipped each month's new feature work, as a share of that month's total. Bands of similar width mean new value is coming from across the team; one band that stays wide means most of it rests on the same person. Named engineers shipped the most Features over the period. Everyone else is grouped as Others.

Engineers and the Features / Maintenance / Tests / Docs / Fixes mix for each quarter. Cost / Perf Unit is what one unit of performance costs against Q2'25. Engineers vs headcount is measured against this org's own pre-AI baseline instead, its first reading in the index, so it agrees with the capacity tile and chart above.