Podcast guest program
Synthetic-voice podcast segments
The Data Nerd records 5, 7, and 10-minute podcast segments via synthetic voice — the same voice that powers our methodology walkthrough. Hosts receive a verbatim 60-second opening, five Q&A prompts, and links to receipts (SSRN paper, methodology, dataset). The format is methodology-first; the founder identity is incidental.
Prepared topics
Seven segments, ready to ship. Each comes with a verbatim opening transcript, host-Q&A questions, and a receipts pack for show notes.
7-min segment
GitHub Commit Velocity Is the Most Leading Public Signal in VC. Here's the Panel.
How a 219-company longitudinal panel showed a median 21-to-47-day lead time over Series A announcements.
- · AI/ML researchers curious about applied alternative-data work
- · Venture capitalists wondering if public data can produce real edge
- · Engineers who want to know what their own commit graph signals to outsiders
7-min segment
Code-Side Sourcing: Naming the Category That Sits Between Crunchbase Scraping and Warm Intros
Why the third era of VC sourcing — reading public engineering activity at scale — needed a name before it could be priced.
- · Emerging-manager GPs building thesis-led funds
- · VC platform teams evaluating which alternative-data layer to buy
- · Founders curious how investors will discover them in the next 24 months
7-min segment
What Agent-Native VC Tooling Actually Looks Like — and Why We Charge Per Call
Free MCP server, EUR 0.19 per deep-signal call via x402 USDC, no subscription gate. Here's why.
- · AI engineers building investor-facing agents
- · VCs experimenting with Claude/Cursor for diligence workflows
- · Anyone curious how x402 micropayments will reshape API pricing
5-min segment
Why a Venture-Capital Product Published Its Methodology on SSRN — And Re-Released the Data CC BY 4.0
Open data is not anti-business; it's the only defensible moat when the next product is shipped by a team three times your size.
- · Open-source maintainers wondering if 'open core' works in VC tooling
- · Researchers curious whether industry data ever makes it into peer review
- · Founders evaluating moat strategies in commoditizing markets
10-min segment
How to Read 4,200 GitHub Organizations a Week Without Getting Rate-Limited Into Oblivion
The systems story behind the panel — five primitives, three layers of caching, one war with the secondary-rate-limit ledger.
- · Senior engineers doing data-pipeline work at scale
- · SREs curious about the GitHub API rate-limit topology
- · Anyone building public-LLM-grounded products at scale
5-min segment
The 38 Percent False-Positive Rate We Won't Hide — And Why That's the Whole Point
Why publishing the failure rate of your signal is the only credible move, and how 'silent rounds' broke our intuition.
- · Data scientists who love a good post-mortem
- · Investors who want a rigorous take on alt-data overclaiming
- · Anyone who's tired of vendor decks claiming '95% accuracy' with no error bars
7-min segment
We Gave Away the MCP Server. Here's What It Did to the Top of the Funnel.
How a free six-tool MCP server became the highest-converting acquisition channel for a paid VC product.
- · Founders deciding whether to launch an MCP integration
- · Indie hackers exploring agent-native distribution
- · AI/ML engineers shipping their own MCP servers and wondering if the funnel works
Active outreach
15 shows in our ready queue. Updated as shows respond.
Hosting a show that fits? Email press@gitdealflow.com with the show URL and audience demographic. We’ll send the verbatim 60-second preview and a synthetic-voice sample within 48 hours.