Sudhir Tiwari

One of the most enjoyable parts of my job as CMO of Thoughtworks is the in-person time we have every quarter as a global management team (GMT). A lot of thought and preparation goes into our meetings to respect the investment of time and money. Timely pre-read, sessions with clear objectives, and prioritising the big bets that underpin our collective success. Its a time to meet the local team, partners and clients. And of course, there is banter, laughter, good dinners, and wine!

Two highlights from the week so far here in San Francisco

1. My first escape-room ‘the glitch’ – a well-tested partnership between the CMO and CTO 🤣 helped crack two of the codes. We escaped with the generosity of a few extra minutes!

2. A full afternoon with the GMT – hands-on keyboards showing agentic concepts, prototypes, and production agents built in the last 30 days. Our CEO Mike Sutcliff– expects us to ‘do, not talk’ about AI. Roll our sleeves up and build.

Amit Choudhary our COO – shared good progress in security agents and agents to assist with deal review

Simone Thompson – Our SVP of partner – showcased an ecosystem partner sales workspace enabling significant growth in our partner-led bookings.

John E. – Global MD of Technology Advisory – demonstrated a strategy/IP vacuum agent to help us be a better strategic partner for our clients.

Shayan Mohanty our CAI and Karthik Srinivasan EVP of ai/works – shared the latest platform upgrades – good progress on open weight models, guard rails and token cost management.

Sudhir Tiwari who leads our global delivery – showed an impactful agent to make program inceptions more impactful.

I had the opportunity to work with our AI engineering team in marketing – Gautam Koul Lucian Moss Rupali R. Tarkesh Deva Natalie Drucker The team has advanced agents deployed in the marketing workflows and has won awards for their work.

In the last 12 days we turned our minds to sales. Sellers spend 30% of their time on admin, and we developed a new agentic prototype to address eight sales motions to attack that 30% with autonomous agents. We moved from concept, UAT to prototype in 12 days. I’m delighted that we have the go-ahead for sprint two.

We move – fast, safely. The prototype included the experience layer, orchestration, semantics, context, intelligence, and data.

We have a federated model at Thoughtworks – clear agentic architecture and guardrails in collaboration with Global IT Services. In sprint one we reused GTM agents, sub-agents, and skills. We designed a prototype in claude design and developed a user video. Rapid UAT came next, and then we built a prototype in python backend, importing the Claude design via MCP into React for the front-end. The team will share more on the technical learnings shortly.

Rachel Laycock Martin Fowler Craig Stanley Steve Yurisich Gene Reznik Ramona Mateiu Erin Cummins Nitin Raina Arijit Bose Kaushik Sarkar John Reid-Dodick

Anyone who’s watched a close match knows how quickly momentum can disappear when play keeps stopping.

It’s much the same when enterprise AI is held back by unreliable systems, awkward integrations or constant troubleshooting.

That’s why we deliver smart, uninterrupted AI so your team can focus on scoring goals, not fixing bugs.

Explore our AI that works approach: https://lnkd.in/d7Q4GvZp

> Based on a conversation I had with Richard Russell.

Watching the World Cup, I keep seeing more than football. VAR, optical trackers, microchipped balls and semi-automated offside calls turn the pitch into a miniature of our AI future where every movement is captured, judged, archived.

That panopticon changes the game.

The old dark arts (dives, shirt pulls, little fouls behind the referee’s back) get harder, but we humans will not stop gaming the system, we will just learn to game the algorithm instead.

But deeper problems are also becoming visible; for example, the offside rule written to stop unfair goal-hanging becomes a sterile tripwire under video surveillance. A toe over the line might as well be ten yards. The letter of the rule is beginning to replace its spirit.

My 2c: machines don’t have to replace us to change us. They reshape our incentives, our behavior, our sense of fairness. It goes well beyond football. In workplaces, schools and markets, people adapt to whatever the machine measures…

Read today’s newsletter if you want to learn more….

Two years ago, I stepped into the role of CEO at Thoughtworks knowing AI would be transformational. But I didn’t expect it to redefine software delivery this quickly.

Yet what gives me the most confidence isn’t the pace of the technology. It’s our people. When we first outlined what AI/works™ could become, the ambition was high.

The response from our teams was simple: “Yes, we can do that.” And they did.

I’m more convinced than ever that the future won’t be defined by the technology itself — but by what people are empowered to achieve with it. This is why I love being part of this industry, you’re never quite sure of what’s coming next.

I’m grateful to our clients, partners and all Thoughtworkers for the incredible journey so far, and I’m excited for what’s next.

More of my reflections in the article below.

Manish Kumar, Market Director, Thoughtworks, joins as a partner speaker at the ETGCCWorld Global Capability Centers Growth Summit 2026, Pune, bringing a sharp perspective on how engineering, AI, and design-led thinking are reshaping enterprise transformation.

Be part of the conversations driving the next wave of innovation-led GCC growth.

Register now: https://lnkd.in/gR7zM8uc

#ETGCCGrowthSummit #ETGCCGrowthSummitPune #GCC #GlobalCapabilityCentres #GCCLeaders #Leadership #GCCCommunity #GCCNews #GCCTrends #ETGCCWorld

It’s been a long, hectic trip: six cities in eight days, crisscrossing the country, and I was exhausted.

To top it all, my cabin bag at security is marked to be opened and checked by the CISF (Central Industrial Security Force) security officer.

However, as he is doing that, he looks up at me and says, “Are you not the person who has done the AI Literacy training course recently?” I perked up. He said he loved it, and lots of his CISF colleagues have taken the course!
I am already feeling better. So is my co-founder, co-author, and co-course creator, Anuj Magazine.

We request the officer for a picture. He obliges, and puts on his hat.

As we are lining up to take the selfie, this other guy in a green t-shirt walks up, looks at me and says, “Hey, I took your AI course. Thanks, I know so much about AI know.” What are the odds??

By now, we are beaming, exhaustion forgotten, and request him to join in the selfie.

We have dedicated the last couple of years of our life to building AI Literacy. In organisations, and beyond.

Nothing has been as satisfying as the Yuva AI for All course, which we built on the request of IndiaAI and Ministry of Electronics and Information Technology.

This is a mission to make every Indian AI Literate

We are led to believe that more than 85 lakh (8.5mn) Indians have registered, and more than 25 lakh have already taken that course, and got a certificate from the Government of India!

We bumped into two of them – both at Airport Security 😊

The course is available free at nasscom FutureSkills Prime, Coursera, Khan Academy , Simplilearn, Tata Consultancy Services iON, Intellipaat, iGOT.AI Karmayogi (for govt. employees).

And, best of all, it is now available in English, Hindi, Kannada, Marathi, Malayalam, Odia, Punjabi, Tamil, Telugu, Assamese, Bengali and Gujarati on TCS iON / Futureskills Prime (link in the comments/reply below)

If you are not yet among the 85-lakh people, go check it out. And spread the word.

A huge thanks to Abhishek Singh Pramath Raj Sinha Lokesh Anand among many others

AI&Beyond Prerna Singh Bindra Archana Muthappa

2025 was the year of AI experimentation. 2026 is the year enterprises must operationalize it.

As organizations deploy more autonomous agents, they’re encountering a familiar challenge at a new scale: governance.
AI agents can access data, invoke tools and execute workflows. But without visibility, controls and oversight, agent sprawl can quickly create security, compliance and cost risks.

That’s why we’re excited to introduce Agent/works™ by Thoughtworks.

Agent/works™ provides a governed runtime and centralized control plane for AI agents across any cloud, helping enterprises manage risk, control AI spend and scale agentic systems with confidence.

We’re showcasing Agent/works™ this week at Databricks Data + AI Summit.

Learn more: https://lnkd.in/djuax6zw