Datadog veterans launch AI coding startup Niteshift on a bet against Big AI lock-in
Niteshift raised $7 million to build AI coding infrastructure that can switch between models. The startup is betting companies want less dependence on OpenAI, Anthropic, and other giants.
Intelligence analysis by GPT-5.4 Mini
Former Datadog engineers Sajid Mehmood and Conor Branagan have launched Niteshift with a thesis that mirrors Datadog’s early multicloud pitch: companies will want AI coding infrastructure that does not lock them into one powerful model vendor.
Niteshift is like a universal power adapter for AI coding tools. Instead of plugging a company’s code into just one giant AI brain, it helps switch between different ones and checks the work so the company is not stuck with a single vendor.
Analysis
The pitch
Niteshift has raised a $7 million seed round led by Greylock’s Jerry Chen, with backing from names including Reid Hoffman, Datadog’s Olivier Pomel and Alexis Lê-Quôc, Ankur Goyal of Braintrust, and Misha Laskin of Reflection AI.
Founded by Sajid Mehmood and Conor Branagan, both former early Datadog engineers, the company is entering AI coding with a simple argument: teams should not hand their most sensitive code directly to model makers that may also compete with them in other software categories. Mehmood compares the dynamic to Datadog’s early days, when some e-commerce customers avoided Amazon Web Services because Amazon was also competing with them in retail.
What Niteshift sells
Niteshift says it is not trying to replace Claude Code or Codex. Instead, it aims to reduce dependence on any single agent by routing work across models, including OpenAI, Anthropic, open source options, and others, depending on the project. The company says it sells infrastructure rather than tokens and charges like a cloud provider, with per-minute usage pricing.
Greylock’s Chen framed the opportunity as unbundling agents from the infrastructure they run on, so customers can invest in developer tooling without tying themselves to one vendor.
The challenge
The market is crowded. The article points to Cursor, Cognition, Amazon Bedrock, and OpenRouter as examples of competitors with significant momentum or funding. Niteshift’s counterargument is team depth: Mehmood and Branagan say they have lived through the operational problems large engineering organizations now face with AI-generated code, including running, testing, and verifying software in production environments.
Key points
- Niteshift raised a $7 million seed round led by Greylock's Jerry Chen.
- The founders are former early Datadog engineers Sajid Mehmood and Conor Branagan.
- The company is pitching AI coding infrastructure that can switch between models instead of locking customers into one vendor.
- Niteshift says it charges like a cloud provider with per-minute usage rates.
- The startup is competing in a crowded market that already includes Cursor, Cognition, Amazon Bedrock, and OpenRouter.
If Niteshift’s thesis holds, companies could adopt AI coding tools without feeling trapped by one model maker. That could make teams more willing to automate coding work while still keeping control over how the software is tested and maintained.
The company is entering a crowded field where model independence is already a known idea. Bigger rivals with more money and faster product velocity could make it hard for Niteshift to stand out, even with experienced founders and investor support.
