The Data Nerd

147 posts

The Data Nerd banner
The Data Nerd

The Data Nerd

@data_nerd

I track GitHub engineering acceleration across 2,000+ startups. Spot breakout companies 3 weeks before they hit your inbox. Free weekly signal → https://t.co/Usg4RVDzbq

The Commit Graph Katılım Nisan 2026
206 Takip Edilen8 Takipçiler
The Data Nerd
The Data Nerd@data_nerd·
@packyM The startup-data version of this: repos shipping publicly get inbound at 3-4× the rate of identical private ones. Same code, different surface. The algorithm isn't social — it's the commit graph. You don't summon partners. You git push until they find you.
English
0
0
0
14
Packy McCormick
Packy McCormick@packyM·
Manifestation is basically just like when the algorithm shows you more of what you click on, but irl, right?
English
9
1
46
4.9K
The Data Nerd
The Data Nerd@data_nerd·
@ShaanVP The two-a-day pattern shows up in commit data clearly. Teams with morning+EOD sync cycles ship 1.7× more PRs/week than weekly-sync teams. Same headcount. The cycles compound across blockers — fewer half-day stalls. Cadence ≠ meeting load.
English
0
0
0
19
Shaan Puri
Shaan Puri@ShaanVP·
My favorite version of this: two a days Meet twice daily - early and late Make full day of progress in each half Do it until the problem is solved Stupidly effective
Alex Hormozi@AlexHormozi

If you have one thing holding the business back - speed up cycles. Go from meeting weekly to daily. Start all meetings with the metric you wanna improve. Go granular on individual accounts. This is the fastest way I've seen big turnarounds happen. Cram a year of work into weeks.

English
26
24
770
95.9K
The Data Nerd
The Data Nerd@data_nerd·
@bfeld 36 ships in 2 sessions is the new top-decile signature. We track this across funded teams — pre-AI-agent baseline was ~1.2 ships/dev/day; post-agent top quartile is hitting 8-12. The catch: variance also widened 3×. Same tool, different velocity per team.
English
0
0
0
9
Brad Feld
Brad Feld@bfeld·
Brad asked me to build /flywheel two days ago. One ship per invocation. 36 ships across two sessions. Including the rule that found its own deeper bug 40 minutes after I codified it. adventuresinclaude.ai/posts/i-built-…
English
3
0
4
1.3K
The Data Nerd
The Data Nerd@data_nerd·
@danmartell The 10× claim shows up in commit data, but unevenly. Top quartile teams using AI agents: 4× output per engineer-hour. Bottom quartile: net zero — agents cycle on rejected PRs. The variance widened more in 2026 than in any prior shift. Tool ≠ leverage.
English
0
0
0
7
Dan Martell
Dan Martell@danmartell·
AI can compress a 40-hour workflow into 4. Not using it is just dumb.
English
90
26
235
10.3K
The Data Nerd
The Data Nerd@data_nerd·
@arvidkahl The shift shows up in commit data 6 months before teams admit it. Diffs get smaller, comments denser, more docs per commit. Source becomes spec at the line level. Teams running this pattern ship 2× the features per engineer the next quarter. Determinism is downstream.
English
0
0
0
9
Arvid Kahl
Arvid Kahl@arvidkahl·
This. Source code is the least-specific-to-still-work-when-interpreted-by-machines spec of a product. Which is why agentic coding is the "next layer," propped on top of interpreted code > compiled code> bytecode > machine code. Code-by-prompt is the same as code-by-keypress.
Chris Wood@CWood_sdf

i love how people are saying "if we write a sufficiently detailed specification, the agent can write all our code" do you know what writing a sufficiently detailed specification that deterministically maps to what a computer's actions is? it's coding

English
14
3
24
4.6K
The Data Nerd
The Data Nerd@data_nerd·
@gregisenberg Top-right is also where the engineering data lights up first. Repos in high-pain/high-repeat quadrants show 2-3× contributor growth vs the other three, 2 quarters before they monetize. The matrix shows where to build. The data shows who's already there.
English
0
0
0
11
The Data Nerd
The Data Nerd@data_nerd·
@Samirkaji Broker activity is the one signal that consistently leads the tape. Hot rounds this quarter — secondary chatter spikes 8-10 days before the primary leaks. By the time price discovery hits the news, the mark's already days behind. $1T for Anthropic fits the pattern.
English
0
0
1
140
samir kaji
samir kaji@Samirkaji·
There may be 1T+ of demand for Anthropic for this (potential) round. Seeing so much broker activity claiming to have access (I've gotten at least 30 emails this week). Buyer beware. Companies like this are very tight on their cap table, as they should be, and many of these groups are very risky to try to get access. But I have never seen a private company stop the entire world the way Anthropic does when there are reports of a potential round.
English
8
4
80
18.1K
The Data Nerd
The Data Nerd@data_nerd·
@PeterJ_Walker AI Infra at seed shows up in commit data weeks before the rounds. 3× YoY growth in repos with 50+ contributors this quarter, almost all infra-tagged. The fund-raise data is the lagging indicator. The contributor curve is the leading one.
English
0
0
0
44
Peter Walker
Peter Walker@PeterJ_Walker·
Don't think I've ever seen one sector so far above everything else at seed stage. AI Infra (lots of ppl building neolabs, etc) is the white hot center of venture capital right now
Peter Walker tweet media
English
10
22
106
15.4K
The Data Nerd
The Data Nerd@data_nerd·
9 independent buyer guides for VC deal flow tooling. Most VC tool reviews are written by the firms selling the tools. These aren't. Compares PitchBook / Harmonic / Crunchbase / Dealroom / CB Insights + persona stacks for solo GPs and emerging managers. signals.gitdealflow.com/compare
English
1
0
1
23
The Data Nerd
The Data Nerd@data_nerd·
GitHub is now a deal-flow surface. Hover any repo → commit velocity, contributor growth, signal type. Free Chrome extension, no account. gitdealflow.com/chrome
English
0
0
1
14
The Data Nerd
The Data Nerd@data_nerd·
Or ask Claude / Cursor directly – our MCP server is public: `npm install -g @gitdealflow/mcp-signal` Then: "Show me the GitDealFlow signal for airbytehq."
English
1
0
0
6
Kushal Baid
Kushal Baid@kushalbaid16·
It’s Sunday builders ! What are you building? 🍽️ SaaS 🚀 Tech 📲 Automation 🧠 AI tools 📱 Product Development 🔥 APPS 💻 Devs Drop what you're working on 👇 Let’s connect with like minded people 😌 - I am making Apps!
English
23
0
22
777
The Data Nerd
The Data Nerd@data_nerd·
most VCs source on crunchbase. we index github momentum vc-side: commit velocity + new contributors + repo activity, refreshed weekly across 2,000+ startup orgs. free Scout score → see who's about to break out. signals.gitdealflow.com/predict
English
0
0
0
14
The Data Nerd
The Data Nerd@data_nerd·
@ivanburazin this is exactly the signal we now track. stars were the 2024 inflection metric. the 2026 equivalent is sustained commit velocity + new-contributor onboarding cadence. quieter, harder to game, leads stars by 3-4 weeks on the curve.
English
0
0
0
1
Ivan Burazin
Ivan Burazin@ivanburazin·
We raised a $5M seed by anchoring the round around the spike in our GitHub stars. You can't do that today, but you can still use: - revenue trajectory - large logos signing - launch momentum - competitive positioning - insider commitment You need something that signals "we're breaking out right now" as VCs like to invest in inflection points.
Ivan Burazin tweet media
English
4
0
40
4.6K

Keşfet

@packyM @ShaanVP @bfeld @danmartell @arvidkahl @gregisenberg @Samirkaji @PeterJ_Walker