
Datadog Veterans Niteshift Raises $7M Seed Round Led by Greylock’s Jerry Chen
Key Takeaways
- Niteshift secures a $7 million seed round led by Greylock's Jerry Chen.
- Founders are two former Datadog engineers.
- Aims to combat Big AI lock-in by enabling cross-provider model swaps.
Niteshift bets on portability
Niteshift, an AI coding agent startup founded by former Datadog engineers Sajid Mehmood and Conor Branagan, raised a $7 million seed round led by Greylock’s Jerry Chen.
“Niteshift raises $7M seed round to challenge Big AI lock-in The AI coding agent startup is betting that enterprises will eventually demand the freedom to swap model providers without rebuilding from scratch”
TechCrunch says Niteshift’s bet is that companies will increasingly want infrastructure that separates the coding model from other orchestration needed to ensure AI-generated code is properly vetted and maintained.

The startup’s strategy is framed as a way to avoid “Big AI lock-in,” with Niteshift routing between GPT and other open-source models based on project needs.
Crypto Briefing similarly describes Niteshift as positioning itself against making it painful for customers to leave, arguing enterprises will want control over which large language models they use rather than being “welded” to a single provider’s ecosystem.
Founders cite Datadog’s lesson
Mehmood compares the current AI coding market to Datadog’s early growth, when e-commerce businesses refused to build on Amazon Web Services because Amazon was their main competitor in retail.
In TechCrunch’s account, Mehmood argues that companies should not trust “their most sensitive assets — code that runs its products — directly to model makers like OpenAI and Anthropic,” because those companies are constantly “killing” startups by launching competing apps.

TechCrunch also quotes Mehmood saying, “Being able to switch between GPT and cloud models is important,” adding, “Everybody’s worried about getting stepped on by these giants.”
The Zamin.uz report says Niteshift’s core question is why companies should entrust their source code directly to model creators like OpenAI or Anthropic, and it says the company aims to reduce dependency on tools like Claude Code or Codex rather than fully replace them.
A crowded market, no tokens
TechCrunch places Niteshift in a crowded AI coding space and lists competitors including Cursor, Cognition, Amazon Bedrock, and OpenRouter, while noting that Niteshift is not selling tokens and instead charges like a cloud provider with per-minute usage rates.
“Niteshift, astartupfounded by formerDatadogengineers, has raised $7 million in a funding round led by Greylock”
The TechCrunch story says Niteshift’s platform routes between Claude Code and Codex along with open-source options and others based on the needs of each project, and it frames the goal as letting customers invest in developer tooling without locking themselves into a single model or agent vendor.
Jerry Chen’s rationale is quoted by TechCrunch as an “opportunity to offer customers an alternate path: unbundling their agents from the infrastructure they run on,” and he says Niteshift is building the platform that enables this for coding agents.
Crypto Briefing adds that Niteshift appears to be taking the traditional SaaS route, building a product that sits between enterprises and model providers rather than using blockchain technology, crypto tokens, or decentralized infrastructure.
More on Technology and Science

Logitech Launches Mobi Fold Travel Mouse Priced at $79.99 With 22% Less Muscle Strain
16 sources compared

Türkiye Unveils COP31 Global Electrification Target: 35% by 2035
16 sources compared

Starlink Adds $10 Monthly Kit Fee for New Users, Replacing One-Time Hardware Rentals
10 sources compared

Researchers Recover Ancient DNA From 700,000-Year-Old Ground Squirrel Droppings in Beringia
11 sources compared