Deepdive: How 10 tech companies choose the next generation of dev tools

2026-02-03 · Source: The Pragmatic Engineer · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering, Project & Product Management · Depth: Advanced, long

Summary

Ten tech companies, ranging from a 5-person seed-stage startup to a 1,500-person public company, are actively re-evaluating their developer tooling stacks, moving beyond GitHub Copilot and ChatGPT. This shift is driven by the emergence of new tools like Cursor, Claude Code, Codex, Gemini CLI, and AI code review tools such as CodeRabbit, Graphite, and Greptile. Small teams (under ~60 engineers) adopt tools quickly based on developer sentiment and organic spread, often through short two-week trials. Larger companies (150+ engineers) face slower adoption due to security reviews, compliance, executive budgetary concerns, and vendor lock-in. A universal challenge across all company sizes is the inability to effectively measure the actual productivity gains from AI tools, with metrics like "lines of code generated" being widely distrusted by engineers and executives alike.

Key takeaway

For VPs of Engineering or Directors of AI/ML evaluating new developer AI tools, recognize that adoption hinges on developer trust and demonstrable value, not just features. Prioritize tools that can navigate your organization's security and compliance frameworks, and be prepared to champion solutions that offer clear, if not always quantitatively measurable, productivity gains, especially when transitioning from established, lower-cost options to more advanced, pricier alternatives like Claude Code.

Key insights

Developer trust and clear value propositions drive AI tooling adoption more than top-down mandates.

Principles

• Tooling decisions scale with company size.
• Security and compliance are major hurdles for large enterprises.
• Directly measurable impact accelerates tool approval.

Method

Small teams use short trial periods and organic adoption. Larger companies require structured evaluations, beta trials, and C-level sponsorship to navigate bureaucracy and budgetary constraints.

In practice

• Conduct short, developer-led trials for small teams.
• Prioritize tools with clear security and compliance roadmaps.
• Use "show and tells" to share tooling setups.

Topics

• AI Developer Tools
• AI Code Review
• Developer Productivity Measurement
• Tool Adoption Strategies
• GitHub Copilot

Best for: VP of Engineering/Data, Director of AI/ML, Software Engineer, Machine Learning Engineer, CTO

Related on AIssential

• AI Tooling for Software Engineers in 2026 — AI tools are now mainstream in software engineering, with Claude Code rapidly dominating usage and preference.
• The Pulse #157: Internal dev tooling at Meta & the “trajectories” feature — Big Tech's internal tooling, monetization strategies, and regulatory challenges are evolving rapidly.
• In an AI world, the most valuable developers will be both artisans and builders — Developers must integrate AI's speed with deep craftsmanship to build robust, secure software in the evolving tech landscape.
• Hype and Reality of the AI Coding Shift — Rapid AI coding adoption creates a critical trust and verification gap, shifting developer toil from creation to quality assurance.
• Domain expertise still wanted: the latest trends in AI-assisted knowledge for developers — Developers increasingly use AI for learning and work, but a significant trust gap necessitates validation with traditional resources.
• How Coinbase scaled AI to 1,000+ engineers | Chintan Turakhia — Large engineering organizations can achieve significant AI adoption and velocity gains through hands-on leadership and strategic tool integration.

Counsel's verdict on this

AIssential's Counsel cites this article in its editorial verdict on the decision it informs:

• Roll AI coding tools across the whole engineering org — and how do we measure it? — Enterprises hold pilot mode for 12–18 months to establish ROI, as unrestricted AI tool usage scales to $3,000 per developer monthly and task-specific model routing yields 10–15% greater developer productivity.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by The Pragmatic Engineer.