Skip to content

Chapter 8: Teams for AI-First Companies — Resources

Curated resources for deeper exploration of topics covered in this chapter.

Frameworks from This Chapter

  • 90-Day AI Fluency Program — Structured three-phase program (Foundation, Integration, Mastery) for building organization-wide AI fluency with measurement frameworks and break-even timelines.

Tools & Platforms

  • GitHub Copilot — AI pair programming assistant; developers using it complete tasks 55% faster and 126% more projects per week (referenced in Section 3: Hiring for AI-First)
  • Cursor — AI-native code editor used by Perplexity engineers as an "invisible coding partner" for scaffolding and boilerplate (referenced in Section 1: The 4 Team Models)
  • Sourcegraph Prompt Library — Platform for saving, sharing, and promoting frequently used prompts across engineering organizations (referenced in Section 6: Building AI Culture)
  • PromptHub — Git-like version control for prompts with branching, merging, and performance tracking (referenced in Section 6: Building AI Culture)
  • CoderPad — Interview platform used by Meta for their AI-enabled coding round with AI chat panel (referenced in Section 3: Hiring for AI-First)
  • Claude Code — Anthropic's AI coding assistant; Caddi CEO encourages candidates to use it during technical exercises (referenced in Section 3: Hiring for AI-First)
  • Uber Michelangelo — Uber's ML platform standardizing data processing, experimentation, and model deployment across 400+ active ML projects (referenced in Sections 1 and 2)
  • Netflix Metaflow — Open-source ML framework enabling hundreds of domain teams to maintain independent ML projects with minimal overhead (referenced in Section 2: Own Your Domain, Share Your Foundation)
  • IBM AI Fairness 360 — Bias detection toolkit (referenced in Section 5: How Roles Are Changing)
  • Microsoft Fairlearn — Systematic bias detection tool (referenced in Section 5: How Roles Are Changing)
  • Figma AI Plugins — AI plugins that generate components matching existing design systems and convert designs into code (referenced in Section 5: How Roles Are Changing)
  • Hugging Face — Platform used by AI-first startups to evaluate candidate portfolios and side projects (referenced in Section 3: Hiring for AI-First)

Further Reading

  • McKinsey: Cultural Factors in AI Failure — Research finding 70% of failed AI initiatives attributable to cultural factors rather than technical limitations
  • HackerRank 2025 Developer Skills Report — Research on AI-enabled personalized tutoring and evolving developer skill requirements
  • Qodo State of AI Code Quality 2025 — Report finding 81% of teams using AI for code review reported quality improvements, but only 3.8% report high confidence shipping AI code without human review
  • Fortune/Harris Poll 2025 — Survey finding 1 in 3 workers actively sabotaged their company's AI rollout
  • Microsoft Work Trend Index — Finding that 71% of business leaders prefer AI skills over experience in candidates
  • Riseworks AI Talent Salary Report 2025 — AI-related roles earning 67% higher salaries on average per Glassdoor data
  • PWC AI Jobs Barometer 2024 — Job postings requiring prompt engineering skills increased 434% from 2023 to 2024
  • Adaptavist Study 2025 — Finding 35% of workers actively gatekeep knowledge to protect job security

Research & Data

  • MIT/Fortune Report — 95% of generative AI pilots at companies fail to influence profit and loss; the core issue is the "learning gap" for both tools and organizations
  • METR Developer Productivity Study — Experienced developers took 19% longer on mature open-source projects when using AI tools, despite feeling 20% faster
  • Anthropic Internal Research — Engineers use Claude in 59% of work (up from 28%); 67% increase in merged PRs per engineer per day; 27% of AI-assisted work "wouldn't have been done otherwise"
  • Scale AI Workforce Data — 12% of contributors hold PhDs, over 40% have advanced degrees; labeling accuracy improved 35% over competitors through technology and workforce development
  • EY Survey 2025 — Companies miss up to 40% of AI productivity gains due to gaps in talent strategy; only 12% of employees receive sufficient AI training
  • S&P Global Data — 42% of AI initiatives scrapped in 2025, up from 17% the previous year

Community & Learning

  • Anthropic's "Member of Technical Staff" Model — Flat structure where every technical employee shares the same title, blurring the research-engineering boundary
  • Meta AI-Enabled Coding Interview — Launched October 2025; 60-minute sessions in CoderPad with GPT-4o mini, Claude 3.5 Haiku, or Llama 4 Maverick
  • JPMorgan GenAI Training Program — Mandated GenAI training for every new employee starting 2024
  • Citi Prompt Writing Upskilling — Began upskilling most of its workforce in prompt writing, September 2025
  • AI Hackathons — One to two-day events building prototypes around real business problems; serve skill building, idea generation, and cultural normalization simultaneously

Companies Referenced in This Chapter

Anthropic, Scale AI, Booking.com, Airbnb, Uber, Siemens, Meta, Perplexity, Netflix, GitLab, OpenAI, Yirifi, Microsoft, Deloitte, Air Canada, JPMorgan, Citi