How Thoughtworks scales AI adoption to empower their workforce

Technology is the easy part. The people? That’s where the real challenge—and the real magic—happens.

At Google Cloud Next 2026, Jeff Deverter (Field CTO at Pythian) shared the stage with Sara Michelazzo (Director, AI Enablement & Organizational Intelligence at Thoughtworks), to discuss how to scale AI adoption. Sara shared Thoughtworks experience, rolling out AI to over 10,000 employees across 18 countries and how they have successfully integrated AI into their business operations.

If you are struggling to get your team to move past playing with toys and into driving real ROI, the Thoughtworks playbook offers five repeatable steps to transform your culture.

Innovation lives at the edge of chaos

Thoughtworks embraces a decentralized culture. Instead of top-down mandates, they encourage different departments to experiment simultaneously. As Sara puts it, at the edge of the chaos is where innovation happens. The goal isn't to control every experiment, but to try a bunch of things, keep what sticks, and discard the rest. This reduces the friction of change by allowing employees to find their own sensible defaults—better ways of working that feel natural rather than forced.

Crowdsourcing knowledge

How do you stay updated when tech changes daily? You crowdsource the learning experience.

Thoughtworks hosts an annual Data and AI event where everyone—from fresh graduates to C-level executives—can demo their work, share prompts, and show off gems they’ve created.

• The Result: A massive, recorded database of peer-to-peer learning.
• The Impact: It levels the playing field. It’s no longer about the smart people in IT teaching the rest; it’s a communal effort that rewards curiosity over hierarchy.

Move beyond IT: The all-function hackathon

Thoughtworks doesn't limit hackathons to developers. They include marketing, sales, and finance.

• The Strategy: Business leaders act as product owners, submitting real-world use cases (like HR staffing or legal compliance).
• The Outcome: One of their most successful tools—a multi-agent talent marketplace—was born from a hackathon. It now uses AI to match 10,000 consultants to the perfect projects based on their profiles, a task previously managed entirely by humans.

Measurable impacts: Saving 1M hours

Thoughtworks estimates they have saved over 1,000,000 hours through AI. Here’s how that breaks down:

• Individual productivity: Replacing hours of meeting prep and document reading with AI summaries.
• Enterprise search: Reducing the 3–5 hours employees spent "finding things" down to mere minutes.
• IT support: 10% of their 20,000 annual support tickets are now "no touch," solved instantly by AI.
• Operational efficiency: Automating license rotation and laptop replacement saved hundreds of thousands of dollars in overhead.

Building the enterprise brain

The holy grail of their AI journey is agentic content modernization. In a company with 30 years of knowledge, data can get messy. Thoughtworks uses autonomous agents to:

1. Scan the knowledge base for duplicates.
2. Remove outdated content.
3. Identify high-value gems to surface for consultants.

This creates an enterprise brain that ensures a junior consultant and a principal partner start their proposals from the same high-quality, sanitized baseline.

Conclusion: Overcoming the bottleneck

When asked for a final piece of advice, Sara shared a humbling quote from her CDIO: "We are the bottleneck."

The technology is ready. The tools are there. The only thing standing in the way of a more productive, more satisfied workforce is our willingness to "unlearn" the old ways and embrace the tension of innovation.

Fast-track months of strategic conversations—clearly define your data and AI strategy.