Niteshift Two Former Datadog Engineers Just Raised $7M to Bet Against AI Coding Lock-In
Niteshift launched today with $7 million in seed funding to build model-agnostic cloud infrastructure for AI coding agents. Co-founders Sajid Mehmood and Conor Branagan spent a decade at Datadog — and are making the same bet their former employer made against AWS.
A startup called Niteshift launched today with $7 million in seed funding, a product that runs AI coding agents in the cloud, and a very specific pitch: the AI vendors selling you coding agents have a conflict of interest in running the infrastructure those agents live on.
The round was led by Greylock’s Jerry Chen. Notable angels include Reid Hoffman, Datadog co-founders Olivier Pomel and Alexis Lê-Quôc, Ankur Goyal of Braintrust, and Misha Laskin of Reflection AI.
Niteshift was founded by Sajid Mehmood and Conor Branagan. Both spent most of the last decade at Datadog, building infrastructure and developer tooling. That background is doing visible work in how they’re framing the company.
“Everybody else is selling labor replacement intelligence,” Mehmood told TechCrunch. “We’re selling software to agents, as opposed to humans — but we’re still out here selling software.”
The Lock-In Argument
The core thesis is familiar to anyone who watched Datadog’s early growth. Enterprise companies that built on AWS in the early 2010s found themselves locked into a vendor with growing ambitions in their own markets. Datadog made a business out of the fact that those customers wanted observability tooling from someone who wasn’t also their compute provider and marketplace competitor.
Niteshift is making the same bet, one layer up. OpenAI sells Codex. Anthropic sells Claude Code. GitHub (Microsoft) sells Copilot. Each of those vendors has a financial incentive to route agent work through their own model and their own infrastructure. Niteshift’s argument is that enterprises will want someone in the middle who isn’t also selling the agents.
“The opportunity is unbundling their agents from the infrastructure they run on,” Chen said, describing why Greylock led the round. Customers can invest deeply in developer tooling without locking themselves into a single model or agent vendor.
What the Product Does
Niteshift is a cloud platform that runs AI coding agents inside fully configured development environments. You connect your repository, define your stack, and Niteshift handles the rest: detecting configuration from CI files and Docker setups, spinning up databases and auth layers, seeding data, and running agents inside isolated environments.
The agents are not Niteshift’s. You bring Claude Code, Codex, OpenCode, or others. Niteshift runs them. When an agent completes a task, it attaches evidence to the pull request: test results, browser checks, logs. The team can review proof that the change actually works before merging.
Pricing follows a cloud model. You pay per minute of compute, not per seat or per token. Niteshift’s revenue doesn’t go up when agents make more API calls to Anthropic or OpenAI — a structural separation the company is emphasizing in early conversations with enterprise prospects.
Why Now
The timing reflects something real. A year ago, running multiple AI agents in parallel meant juggling local worktrees, burning through your laptop’s RAM, and hoping your local environment matched production closely enough to matter. Tools like Codex Remote, Claude Code cloud sessions, and Copilot cloud agents have normalized the idea of agents running in the cloud — but all of them bundle the agent with the infrastructure.
Niteshift is the bet that those will come apart, and that the infrastructure layer will be worth something as a standalone business.
The company is targeting engineering teams already running agents at scale — the ones where the question isn’t whether to use AI coding agents but how to manage 20 of them running at once across 15 repositories.
Source: TechCrunch