Videos
Interview: Kognitos CEO Binny Gill at OPEX Week
Binny Gill at OPEX Week on why English-as-code, not no-code, is the breakthrough that finally lets business users build and maintain enterprise automation.
About this interview
Founder and CEO Binny Gill speaks at OPEX Week: Business Transformation World Summit about why training machines to speak English — not training humans to code — is the genuine breakthrough in enterprise automation.
Three ideas Binny argues for
- Fifty years of training humans hasn't worked: from procedural languages through no-code and low-code, the size of the developer pool has never been enough. IT has always been the bottleneck.
- Train the machine to speak human: Kognitos builds a runtime that faithfully executes natural-language business instructions the same way machine code is faithfully executed — without translating them into developer-shaped artefacts first.
- RPA still needs developers — and breaks anyway: when a business hands a task to an RPA team, the developer maps it into a workflow tool. When an undocumented edge case appears, the IT team has to come back. Kognitos avoids that hand-off by keeping the source-of-truth in the English the business user wrote.
Why OPEX matters as a venue
Binny notes that OPEX Week attracts the exact buyer profile Kognitos serves: operations leaders from different industries who are actively curious about generative AI and looking to apply it to specific pain points. The conversations validated that the problems Kognitos is solving are the ones operations teams are already trying to solve themselves.
Questions answered in this video
What does Kognitos's CEO mean by “training machines to speak human”?
Instead of building yet another tool for developers (or yet another low-code workflow builder), Kognitos built a runtime that takes natural-language business instructions and executes them faithfully — the same way a CPU executes machine code. Business users author and own the program in English.
How is Kognitos different from RPA according to Binny?
RPA still requires a developer or developer-mindset person to translate a business task (originally expressed in natural language) into a workflow. When an unmodelled edge case appears, IT has to come back to fix it. Kognitos keeps the English description as the executable source of truth, so business users own both the original automation and its evolution.
Who is the right audience for Kognitos according to the OPEX conversations?
Operations leaders — finance, supply chain, customer service, healthcare operations — who are actively curious about generative AI and looking for specific business pain points to address with it. Binny said the OPEX audience matched Kognitos's buyer profile almost exactly.
What does Binny see as the structural problem with five decades of automation?
The industry has tried to widen the pool of people who can speak the language of machines — through procedural languages, no-code, and low-code — and still IT teams are overburdened. The fix isn't to train more humans to code; it's to train machines to understand and execute natural language.