About
Dr. Hitesh Sarda has built production software for 25 years, from a global bank in Frankfurt to startups in Hyderabad, and now advises leadership teams on AI.
I am Dr. Hitesh Sarda. I advise leadership teams on AI: what to build, what to buy, what to skip, and how to prove a use case with a small pilot before anyone commits a budget. I run 256 Technologies, an AI consulting and advisory firm and applied AI lab in Hyderabad, and I wrote Implementing AI Agents in Your Company, a free book for the managers who have to make this technology pay.
I came to that work the long way, through 25 years of building software that had to run.
How I got here
It started with a computer my father could barely afford. He bought me an HCL Beanstalk PC during a difficult stretch for our family, and as I wrote when I turned 42, I owe my career to him.
After an MBA I was placed from campus into the technology arm of a large global bank, and that job set the direction for everything after. It took me to Frankfurt for seven years, where I learnt how to build, run and support enterprise-scale systems, working up and down the stack long before “full stack” was a job title.
I have also done the opposite kind of work. I co-founded startups twice. At INSORCE, which sold technology to global financial institutions, I was the technical co-founder and CTO, which meant budgets, hiring, firing, architecture, product management and writing code in the same week. In 2014 I gave that job a name, the Full Stack CTO, and it is still my handle on X.
Since then I have worked across banking, payments, automation and retail, including production systems work at Citi, Virtusa, Yash and ValueLabs, and loyalty and marketing technology at OptCulture, where I am now an advisor.
Early to the hard parts
I have a habit of getting interested in a technology before it is fashionable and writing down what I find.
- 2009: started this blog, and have kept it going since.
- 2011: finished Stanford’s online machine learning class and published my notes.
- 2017: was delivering robotic process automation to customers, and argued that RPA is duct tape: useful, and no substitute for fixing the process. Spoke on a blockchain panel at Fintech Valley Vizag with HDFC Life, SBI and KPMG.
- 2018 to 2020: wrote on GDPR, ISO 20022 and banking regulation, the unglamorous rules that decide what software in finance is allowed to do.
- 2026: profiled in ET Focus, Hyderabad, on India’s AI opportunity, and published the book on putting AI agents to work in a company.
That pattern matters to the people I advise. I have watched several waves of technology arrive with the same promises, and I know what the gap between a demo and a production system costs, because I have worked on both sides of it.
What I do now
Most of my work is helping a leadership team settle one AI decision at a time. It usually takes one of four forms:
- AI advisory, starting with a single advisory hour.
- Fractional Head of AI, for companies that need senior AI judgment a few days a month.
- AI training for leadership teams, a six-week program built around the team’s own work.
- Speaking, at leadership offsites, industry events and engineering organizations. Recent talks include AWS ISV Ignite, IIM Udaipur and the Visakhapatnam Retail Summit.
In 2026 I completed a Doctor of Business Administration at IIBM Institute and SIMA Switzerland. My dissertation studied how 371 B2B SaaS companies price and package AI features, which is the question behind much of my advice on what AI really costs.
My deepest experience is in banking, retail and loyalty. I am based in Hyderabad and work in person there, and remotely with teams across India and abroad. There is more on that on the AI advisor in Hyderabad page, and the company site has the AI consulting services in Hyderabad in full.
What I believe
A few ideas run through everything I write and every engagement:
- AI adoption is mostly an organizational problem, and the technology is the easier half.
- Most companies should buy before they build.
- An AI pilot exists to reduce uncertainty about the business, so design it to tell you the truth.
- The model is a commodity. Your data, your workflow and how fast your company learns are where the advantage is.
- Reward people for saying “I don’t know”. Confident wrong answers are what AI makes cheap.
The guides collect the arguments behind these by theme: what AI really costs, deploying AI agents safely, leading teams that code with AI, AI strategy for leadership teams and AI in retail and loyalty.
In the press
The Hans India, 10 March 2017. "Fintech Valley Spring Conference 2017 surpasses all expectations". Conference report naming Hitesh, then a Director at ValueLabs, on the panel on how blockchain is transforming financial services infrastructure, alongside HDFC Life, SBI, KPMG, Angel Broking and TenX. About the panel.
Away from work
I live in Hyderabad with my family. I sketch when I can, and this year I designed our Ganpati wall with AI and had it laser cut.
Get in touch
To talk through a specific problem, book an AI advisory hour. To have me speak, use the speaking page. For anything else, email or LinkedIn is best, and I am on X as @fullstackcto.