AI Consulting for Companies

Thoughtworks for AI consulting: engineering-led agents and AI strategy

A software engineering consultancy that builds production agents and plans for vendor independence. Best for companies with engineers of their own.

Contents
  1. What Thoughtworks does for a company
  2. Published evidence
  3. Strengths and limits
  4. Pricing
  5. Questions to ask Thoughtworks
  6. Sources
TypeEngineering consultancy
HeadquartersChicago, Illinois, US
Founded1993
PricingNot published
Websitethoughtworks.com

Thoughtworks is a software consultancy headquartered in Chicago that listed on Nasdaq in 2021 and was taken private again by Apax Partners in November 2024⁠1. Its AI services focus on getting agents into production and keeping control of their cost⁠2.

Best for: Companies with their own software teams that want production-grade agents and a plan to reduce dependence on any one AI vendor.

Not the right fit for: Companies without internal engineers to take over the system, or buyers who mainly want business-process advice.

What Thoughtworks does for a company

It offers four services: an AI strategy that assesses maturity, reduces vendor dependency and controls token costs; agentic AI built on its AI/works platform; an AI Factory that helps a company build its own AI operating model; and data modernisation so that agents can work with business context and security boundaries. Agent/works is a governance layer that observes and cost-manages agents across models, frameworks and clouds. Partners include AWS, Google Cloud, Microsoft, Databricks and NVIDIA⁠2.

Beyond Médecins Sans Frontières, Thoughtworks reports a 90% reduction in manual data search at an unnamed life-sciences company and a manufacturing prediction model that reached 95% target precision in under five weeks⁠2.

Published evidence

  • Médecins Sans Frontières. Work on Antibiogo, an AI app for antibiotic-resistance testing. Thoughtworks reports 60% faster build times, supporting scale-up across more than 100 labs.⁠2

Strengths and limits

Strengths

  • Addresses model-vendor dependency and token costs directly in its strategy offer.
  • A governance layer (Agent/works) that monitors and cost-manages agents across models and clouds.
  • Publishes named and quantified client work.

Limits

  • Some reported results do not name the client.
  • Pricing and engagement terms are not published.
  • Best suited to companies that will run and extend the software themselves.

Choose another option if: You have no engineering team to hand the system to, or the main need is redesigning back-office work rather than building software.

Pricing

Not published.

Questions to ask Thoughtworks

  • Who on our side would own the system and the Agent/works configuration after handover?
  • How does the AI-readiness assessment end, and what does it cost?
  • How do you measure and cap token costs in production?
  • Which clients of our size can we speak with?

See how Thoughtworks compares with the other firms in the main guide.

Sources

  1. Thoughtworks. Wikipedia. Accessed 8 Nov 2026.
  2. Enterprise AI services. Thoughtworks. Accessed 8 Nov 2026.