BlockBeats reported on June 13 that Niteshift, an AI coding agent startup, has completed a $7 million seed financing round. The round was led by Greylock partner Jerry Chen, and the proceeds will be used to expand product development and market promotion. While the size of the financing is not especially large by current AI industry standards, the investor roster gives the company a notable profile among enterprise software and AI infrastructure circles.
Seed round backed by well-known technology founders
Participants in the round include LinkedIn co-founder Reid Hoffman, Datadog co-founders Olivier Pomel and Alexis Lê-Quôc, Braintrust founder Ankur Goyal, and Reflection AI co-founder Misha Laskin. The group spans social networking, cloud monitoring, AI infrastructure and model development, all areas that connect with Niteshift’s focus on software development workflows for enterprises.
Niteshift was founded by Sajid Mehmood and Conor Branagan, both former early engineers at Datadog. According to the company’s description, the two were involved in helping Datadog grow from a startup into a software company with a market value of tens of billions of dollars. Their new company is focused on AI coding agents and the infrastructure enterprises need when they adopt code generation, code management and automated development tools.
An independent layer between enterprises and model providers
Niteshift’s thesis is tied to the expansion of foundation model companies such as OpenAI and Anthropic into vertical software markets. The company argues that more enterprises will be concerned about handing their core code assets entirely to such potential competitors. Mehmood compared the trend with the reservations early e-commerce companies had about Amazon Web Services, and described the AI industry as entering a “SaaSpocalypse” stage similar to the “Retail Apocalypse.”
In terms of product positioning, Niteshift is not trying to replace mainstream AI coding tools such as Claude Code or Codex. Instead, the company wants to serve as an independent infrastructure layer that connects enterprises with model suppliers. Its approach is to let companies use code generation models while keeping code review, testing, permission management and operations workflows separate from the model layer itself.
Reducing reliance on a single model vendor
Niteshift is betting that enterprises will prefer to split different parts of the software development chain rather than bind code generation, review, testing, permissions and operations to one model provider. In the company’s view, this separation can reduce dependence on a single supplier and better align with the way enterprises manage core code assets.
The company also believes that as AI software development becomes part of core enterprise productivity, neutral platforms that do not directly compete with customers’ businesses will face rising demand. The original BlockBeats item also included a link to join the Beating Feishu AI news channel, which monitors global AI topics and news around the clock.