The archive · Developer & Business Tools · Strategic decision · 2026
Niteshift raises $7M for multi-model coding cloud
Datadog veterans founded Niteshift in 2026 with $7M seed to build neutral coding infrastructure, routing tasks between models.
Niteshift
What the business is
Neutral infrastructure/cloud platform for AI coding agents: routes coding tasks to different large models (OpenAI, Anthropic, open-source models), bills by the minute, and lets companies use AI to write code without depending on a single model vendor.
Starting capital：$7M seed round (led by Jerry Chen of Greylock)
How it started
Sajid Mehmood and Conor Branagan were early engineers at Datadog, taking the company from startup to a multi-billion-dollar valuation. In 2026 they founded Niteshift, starting from an analogy: early Datadog customers in e-commerce refused to run monitoring on AWS because Amazon was simultaneously 'killing' them with its retail business; Mehmood believes the AI version of the 'retail apocalypse' has already begun—Anthropic and OpenAI are rapidly entering vertical software markets such as legal, healthcare, and finance (the so-called SaaSpocalypse), and companies will not entrust their most sensitive production code directly to model vendors with competitive agendas.
What happened
Niteshift entered the crowded AI coding space with a $7 million seed round, led by Jerry Chen of Greylock, with angel investors including Reid Hoffman, Datadog's Olivier Pomel and Alexis Lê-Quôc, Braintrust's Ankur Goyal, and Reflection AI's Misha Laskin. The product is an AI coding cloud: routing among GPT, Claude, and open-source models according to project needs, selling infrastructure rather than tokens, billed by the minute. Greylock's investment logic is that frontier labs are moving up into the application layer, and customers need an alternative path that 'decouples agents from the infrastructure on which they run.'
How it ended up
Still running: as of June 2026, the company has just completed its seed round and has not disclosed customers or revenue; facing competitors such as Cursor, Cognition, OpenRouter, and Amazon Bedrock, Mehmood is betting that the founding team's experience scaling Datadog is the moat.
Background
Niteshift, founded in 2026 by Datadog engineers Sajid Mehmood and Conor Branagan, raised $7M seed led by Greylock's Jerry Chen, with angels Reid Hoffman and Datadog co-founders. It helps enterprises use AI for coding without vendor lock-in.
The team's analogy: e-commerce firms avoided AWS because Amazon competed; now OpenAI and Anthropic enter vertical software, so firms won't trust them with code. Niteshift routes tasks among GPT, Claude, and open-source models, not replacing Claude Code or Codex.
Niteshift sells infrastructure billed by minute, not tokens. Competitors include Cursor, Cognition, Amazon Bedrock, OpenRouter. Mehmood cites team experience scaling Datadog as key advantage.
What has to be true
- Companies do not want to hand their most sensitive production code to model vendors that are also competitors; this is a replay of the cloud-era 'reject AWS' playbook.
- Frontier labs are moving up into vertical software application layers, and conflicts of interest will only intensify, creating real demand for a neutral routing layer.
- Selling infrastructure and billing by the minute rather than selling tokens lets Niteshift avoid head-to-head price wars with model vendors.
- The founding team has experience taking Datadog from early days to a multi-billion-dollar valuation, and the investor list (Hoffman, Pomel, Lê-Quôc) validates that.
What can be applied
Look for fears from the cloud era: companies feared platforms competing with them; now they'll pay for neutral AI infrastructure, not tokens.
Aftermath
As of June 10, 2026, Niteshift has just completed a $7 million seed round and is in the early product stage: the platform is designed to route coding tasks among OpenAI, Anthropic, and open-source models, and the company has not disclosed customers, revenue, or a launch timeline. The main risk is that 'model neutrality' is not a new concept, and competitors have larger funding scale and first-mover advantages; the company's current differentiation bet is on the founding team's large-scale engineering experience.
Sources
- Datadog veterans launch AI coding startup Niteshift on a bet against Big AI lock-in
- Niteshift Raises $7M Seed
spotted an error? The archive wants to know.
Your turn
You just read one. Describe what you are building, and see who is betting on the same thing.
Free account · 3 free questions · no card