Datadog veterans launch AI coding startup Niteshift on a bet against Big AI lock-in

2026-06-10 · Source: AI News & Artificial Intelligence | TechCrunch · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering, Emerging Technologies & Innovation · Depth: Intermediate, short

Summary

AI coding agent startup Niteshift has launched with a \$7 million seed round led by Greylock's Jerry Chen, attracting angels like Reid Hoffman and Datadog founders. Founded by former Datadog engineers Sajid Mehmood and Conor Branagan, Niteshift addresses concerns about "Big AI lock-in" by offering an infrastructure layer that unbundles AI coding agents from model makers like OpenAI and Anthropic. Mehmood likens this to the "SaaSpocalypse," where companies avoid building on platforms that could become competitors, similar to e-commerce businesses shunning Amazon Web Services. Niteshift's AI coding cloud will route between models such as GPT and Claude, open-source options, and others, charging per-minute usage rates rather than selling tokens. This approach aims to reduce dependence on single vendors, despite entering a crowded market with competitors like Cursor, Cognition (which recently raised \$1 billion at a \$26 billion valuation), Amazon Bedrock, and OpenRouter (which raised \$113 million at a \$1.3 billion valuation). The founders emphasize their experience scaling Datadog as a key differentiator.

Key takeaway

For Directors of AI/ML evaluating coding agent solutions, you should prioritize platforms that offer model independence to mitigate "Big AI" vendor lock-in risks. Your teams can avoid future "SaaSpocalypse" scenarios by adopting infrastructure that allows switching between models like GPT and Claude, including open-source options. This approach ensures long-term flexibility and control over your most sensitive code assets, preventing reliance on model makers who may become direct competitors.

Key insights

Companies seek AI coding infrastructure that unbundles agents from model makers to avoid vendor lock-in and competition.

Principles

• Avoid vendor lock-in with AI models.
• Separate AI agents from core infrastructure.
• Competition drives unbundling.

Method

Niteshift's AI coding cloud routes between various coding models (GPT, Claude, open source) based on project needs, charging per-minute usage for infrastructure, not tokens.

In practice

• Implement multi-model AI routing.
• Evaluate AI coding infrastructure costs.
• Prioritize vendor-neutral tooling.

Topics

• AI Coding Agents
• Vendor Lock-in
• Multi-cloud AI
• Developer Tooling
• SaaSpocalypse
• Datadog Founders

Best for: CTO, VP of Engineering/Data, AI Architect, Director of AI/ML, Entrepreneur, Investor

Related on AIssential

• Locked, stocked, and losing budget: AI vendor lock-in bites back — AI vendor lock-in and rising costs are challenging enterprise AI adoption, debunking executive assumptions about easy model switching.
• The AI Revolution Will Not Be Monopolized: How open-source beats economies of scale, even for LLMs — Open-source development is posited as a critical force to prevent monopolization of the AI revolution, particularly for LLMs.
• Rules In The Agentic Shipyard: Desired Agents Must Twist-Lock — Enterprises are building internal AI agent frameworks, requiring vendor agents to conform to their architectural standards.
• The Rent-Seeking Trap: Why Your AI Strategy Needs a Hard Fork — AI centralization is an economic capture strategy, not a technical necessity, creating a "rent-seeking trap" for integrators.
• IBM's AI coding 'partner' Bob hits general availability — IBM's Bob AI coding partner leverages diverse LLMs to boost developer productivity across the full software development lifecycle.
• AI Agents and the Fight for Customer Data — Centralized, open data infrastructure is crucial for AI agents, despite vendor data access restrictions.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI News & Artificial Intelligence | TechCrunch.