Blab AI
Niteshift Secures $7 Million Seed Funding to Build Cloud Layer for AI Coding Agents
When the very companies that supply AI models begin to compete with the customers who use them, a new player has emerged to keep the two worlds apart.
Niteshift, founded by former early‑Datadog engineers Sajid Mehmood and Conor Branagan, announced on June 10, 2026 that it had closed a $7 million seed round. The funding was led by Greylock partner Jerry Chen and included high‑profile angels such as Reid Hoffman, Datadog co‑founders Olivier Pomel and Alexis Lê‑Quôc, Ankur Goyal of Braintrust, and Misha Laskin of Reflection AI.
The startup’s premise is simple yet timely: large enterprises are increasingly reluctant to route their most sensitive codebases through commercial AI model makers like Anthropic and OpenAI. According to the founders, those vendors are expanding into vertical software markets and, in the process, competing with the very customers they serve. Mehmood, who serves as CEO, draws a parallel to Datadog’s early days, when e‑commerce firms avoided Amazon Web Services because Amazon was also displacing their businesses.
Rather than replace popular coding agents such as Claude Code or Codex, Niteshift positions itself as a model‑agnostic routing layer. The platform orchestrates between frontier models, open‑source alternatives, and other providers based on project requirements. Unlike competitors that charge token‑based usage, Niteshift bills per‑minute infrastructure fees, framing itself as a cloud provider for AI coding agents rather than a labour‑replacement tool.
“Unbundling coding agents from the infrastructure they run on gives enterprises the freedom to invest in developer tooling without locking into a single model vendor,” Chen said. “That’s the core opportunity for Niteshift.”
The company enters a crowded field that includes Cursor, Cognition, Amazon Bedrock, and OpenRouter. The founders argue that their advantage comes from having built and scaled engineering infrastructure at Datadog through the exact challenges large organisations now face with AI‑generated code.
The seed round gives Niteshift a runway to develop its full‑stack cloud platform, which will enable teams to run, test, and verify AI coding agents in production environments. The model‑agnostic approach is intended to give enterprises flexibility as the AI coding market matures and as vendors compete for vertical dominance.
At present, the company has not disclosed a product launch date. The funding round, however, signals investor confidence in a niche that bridges the gap between AI model providers and enterprise deployment needs.
Niteshift’s success will hinge on its ability to demonstrate reliable routing, cost efficiency, and security for code that remains proprietary to the hiring organization.
In summary, Niteshift has raised $7 million to build a cloud infrastructure layer that can sit beneath any AI coding agent. The startup’s founders, with Datadog experience, argue that enterprises will want control over the underlying infrastructure to avoid lock‑in with model makers. The round, led by Greylock and backed by a who’s‑who of angels, positions Niteshift to compete with existing infrastructure and routing solutions while offering a model‑agnostic, per‑minute billing model. The company’s next milestones will include product development, early customer pilots, and a public launch date.
Niteshift, founded by former early‑Datadog engineers Sajid Mehmood and Conor Branagan, announced on June 10, 2026 that it had closed a $7 million seed round. The funding was led by Greylock partner Jerry Chen and included high‑profile angels such as Reid Hoffman, Datadog co‑founders Olivier Pomel and Alexis Lê‑Quôc, Ankur Goyal of Braintrust, and Misha Laskin of Reflection AI.
The startup’s premise is simple yet timely: large enterprises are increasingly reluctant to route their most sensitive codebases through commercial AI model makers like Anthropic and OpenAI. According to the founders, those vendors are expanding into vertical software markets and, in the process, competing with the very customers they serve. Mehmood, who serves as CEO, draws a parallel to Datadog’s early days, when e‑commerce firms avoided Amazon Web Services because Amazon was also displacing their businesses.
Rather than replace popular coding agents such as Claude Code or Codex, Niteshift positions itself as a model‑agnostic routing layer. The platform orchestrates between frontier models, open‑source alternatives, and other providers based on project requirements. Unlike competitors that charge token‑based usage, Niteshift bills per‑minute infrastructure fees, framing itself as a cloud provider for AI coding agents rather than a labour‑replacement tool.
“Unbundling coding agents from the infrastructure they run on gives enterprises the freedom to invest in developer tooling without locking into a single model vendor,” Chen said. “That’s the core opportunity for Niteshift.”
The company enters a crowded field that includes Cursor, Cognition, Amazon Bedrock, and OpenRouter. The founders argue that their advantage comes from having built and scaled engineering infrastructure at Datadog through the exact challenges large organisations now face with AI‑generated code.
The seed round gives Niteshift a runway to develop its full‑stack cloud platform, which will enable teams to run, test, and verify AI coding agents in production environments. The model‑agnostic approach is intended to give enterprises flexibility as the AI coding market matures and as vendors compete for vertical dominance.
At present, the company has not disclosed a product launch date. The funding round, however, signals investor confidence in a niche that bridges the gap between AI model providers and enterprise deployment needs.
Niteshift’s success will hinge on its ability to demonstrate reliable routing, cost efficiency, and security for code that remains proprietary to the hiring organization.
In summary, Niteshift has raised $7 million to build a cloud infrastructure layer that can sit beneath any AI coding agent. The startup’s founders, with Datadog experience, argue that enterprises will want control over the underlying infrastructure to avoid lock‑in with model makers. The round, led by Greylock and backed by a who’s‑who of angels, positions Niteshift to compete with existing infrastructure and routing solutions while offering a model‑agnostic, per‑minute billing model. The company’s next milestones will include product development, early customer pilots, and a public launch date.
