The people maintaining crypto
A monthly view of verified internal employees, maintainers, lead developers, team growth, retention, and movement across the ecosystem—derived from recorded GitHub work history.
Complete through 2026-07-01 · People with repeated recorded write authority at an organization; maintainers are internal contributors who merge pull requests.
-22.7% over 12 months
Internal people who merge pull requests
Active now with a merge in the trailing three months
Organizations with verified internal activity
427 joined · 2.1K exited
Internal developer population
People are counted once in their primary organization for a month. Maintainers are the internal employees who merge pull requests; active leads are active now and merged in the current or preceding two months.
Depth of participation
Monthly internal people grouped by the number of distinct days on which recorded work occurred. This is activity depth, not seniority.
Joining, exiting, and moving
Confirmed changes in primary organization. Movement is shown separately from people entering or leaving the measured ecosystem.
Retention by starting cohort
The share of internal people still active 3, 6, and 12 months after their first observed internal month. Cohorts below 20 people are not published.
| Started | People | Month 3 | Month 6 | Month 12 |
|---|---|---|---|---|
| Jul 2025 | 893 | 45% | 33% | 23% |
| Jun 2025 | 789 | 46% | 35% | 25% |
| May 2025 | 889 | 47% | 36% | 26% |
| Apr 2025 | 968 | 46% | 40% | 29% |
| Mar 2025 | 864 | 41% | 35% | 25% |
| Feb 2025 | 802 | 42% | 33% | 26% |
| Jan 2025 | 823 | 48% | 36% | 24% |
| Dec 2024 | 706 | 45% | 37% | 26% |
| Nov 2024 | 732 | 42% | 37% | 27% |
| Oct 2024 | 913 | 45% | 37% | 27% |
| Sep 2024 | 933 | 44% | 38% | 26% |
| Aug 2024 | 902 | 46% | 42% | 31% |
| Jul 2024 | 908 | 45% | 36% | 29% |
| Jun 2024 | 955 | 43% | 35% | 28% |
| May 2024 | 1K | 46% | 38% | 29% |
| Apr 2024 | 962 | 44% | 36% | 29% |
| Mar 2024 | 1K | 43% | 36% | 27% |
| Feb 2024 | 882 | 46% | 37% | 32% |
Maintainer leverage
How many distinct internal authors a maintainer merged code for in the complete month—not a claim about title or seniority.
1median internal authors supported
Organizations building now
Ranked by verified internal people in the latest complete month. Twelve-month change, joins, and exits make growth interpretable.
| Organization | Internal people | Maintainers | 12m change | Joined / exited | 12m commits |
|---|---|---|---|---|---|
| Amazon Web Services | 355 | 331 | -22 | +609 / −693 | 43K |
| pytorch | 294 | 150 | +141 | +637 / −423 | 70K |
| MetaMask | 155 | 155 | +2 | +95 / −96 | 53K |
| OpenAI | 133 | 95 | +93 | +341 / −180 | 25K |
| Hugging Face | 123 | 118 | -25 | +125 / −164 | 37K |
| Open Edge Platform | 121 | 120 | -71 | +105 / −177 | 17K |
| sgl-project | 119 | 90 | +88 | +157 / −52 | 15K |
| Chainlink | 118 | 118 | -17 | +120 / −129 | 25K |
| Juspay Technologies | 113 | 95 | +11 | +134 / −127 | 24K |
| LangChain | 99 | 92 | +43 | +131 / −91 | 25K |
| Ledger | 96 | 91 | +13 | +96 / −81 | 30K |
| Kong | 93 | 85 | +9 | +119 / −114 | 13K |
| Parity Technologies | 92 | 90 | +23 | +102 / −68 | 22K |
| LF Decentralized Trust | 91 | 89 | +1 | +86 / −97 | 13K |
| OpenVINO™ Toolkit | 84 | 83 | +11 | +104 / −105 | 3.6K |
| Stripe | 78 | 76 | +10 | +170 / −154 | 12K |
| Nous Research | 75 | 15 | +68 | +85 / −14 | 8.7K |
| IOHK | 74 | 63 | -8 | +112 / −127 | 16K |
| Meta PyTorch | 73 | 20 | -20 | +280 / −274 | 16K |
| Mysten Labs | 70 | 68 | +2 | +67 / −63 | 20K |
| tensorflow | 70 | 18 | +21 | +191 / −218 | 3.3K |
| Weights & Biases | 68 | 60 | +11 | +113 / −97 | 13K |
| Software Mansion | 63 | 63 | -3 | +51 / −48 | 16K |
| SatoshiLabs Group | 61 | 51 | +1 | +36 / −40 | 18K |
| BitGo | 60 | 60 | -9 | +138 / −131 | 7.4K |
| Block Open Source | 59 | 58 | +38 | +110 / −85 | 12K |
| paddlepaddle | 58 | 52 | -65 | +119 / −159 | 4.4K |
| Worldcoin | 54 | 53 | +4 | +79 / −63 | 11K |
| ggml | 54 | 32 | +17 | +99 / −78 | 4.4K |
| Adyen | 51 | 49 | +10 | +59 / −53 | 14K |
Largest observed team flows
Top organization-to-organization moves in the latest complete period. These are inferred from changes in primary internal affiliation, not profile claims.