OKRA straight to series A: What Separates a great machine-learning business from the rest: making machine-learning predictable

During the past few months, I’ve been fortunate enough to share the vision of OKRA with a great community of innovators, clients, venture capitalists and machine learning (ML) experts. Whenever I meet people talking passionately about ML, I always try to understand exactly what it means to them, and why they are so passionate about it. To be honest, most of the time when people talk about ML, they talk about the impact of a solution they believe can be delivered. When I ask, “what problem are you facing?” they don’t always have a clear answer. Typical responses range from “we’d like to use machine-learning because we want to reach out to more patients in need of our drug” or “we want to build great brands for our customers.”


For many questions, the answers are unfortunately not in the data we have today. Even more intriguing, some of our open and honest clients share their experiences on how their internal teams provided them with insights that finally led to poor outcomes and a waste of resources. This has led to increased skepticism, and for clients to sometimes go against the insights provided by data and to follow their gut instincts and ask their customers directly instead.


Machine-learning is about probabilistic learning, and learning is about being humble, reflective, self-aware and honest. In search of the truth, one must dare to acknowledge when one is not confident (i.e. low probability) based on the available data, but also dare to provide with confidence (i.e. high probability) the answer when enough evidence is available. Getting this right is a skill that requires deep knowledge of machine-learning and a high awareness of domain evidence (see my blog about how to learn like a machine). This is our struggle. Our biggest challenge as a truthful machine-learning company with this expertise is to cut through the noise and hype around ML and to avoid disillusionment and fatigue that we see developing with potential clients. A big shift in the mindset and direction of travel of our clients is required to allow for productive use of machine-learning in building evidence-based decision-making into businesses. It’s making it predictable.

I recently raised funding – straight to series A - as I skipped a seed investment round. Most companies that follow this route are typically entrepreneurs with long-standing relationships with venture capitalists or serial entrepreneurs in less experimental high growth markets. Honestly, I am none of above. What I do have however, is the technical expertise to set the bar for the standard of evidence that must be delivered to businesses using machine-learning. I’m extremely proud of how much we’ve achieved in the last two years, building a great evidence-driven artificial intelligence (AI) technology, but most importantly we have built an exceptional team. Although small, we have unique, deep AI expertise and we believe strongly in evidence generation from data. 

When our investor-to-be, a leading entrepreneur and an outstanding social worker with a global, multi-million pound foster care business, heard about this young ‘lady’ building the most trustworthy AI engine to support businesses, he sent his team to meet us at the IET Savoy in London last November on a special mission: Can OKRA build evidence-based foster care in real-time for his global businesses? I asked where did you hear about us? It was amazing to hear that one of our most engaged clients, Rasim Shah, who believed in our cutting-edge technology had referred them on. Fast forward five months and we have also convinced Rasim to join OKRA as CRO to drive our sales and marketing efforts. We truly couldn’t have a better person for the CRO role than a previous client who believes in our technology and why OKRA is on this journey.


Our expert team is the key to our success, respectfully direct and courageous in building machines that deliver the truth, but the team also has the courage to tell you when they don’t know the answer. Our clients are not looking for a system that is too good to be true, but they are looking for the truth. To deliver the best for their customers, patients and brands, it is a platform that delivers the truth that will be of most value in today's world.


OKRA believes that machine-learning is the best tool for evidence-based decision-making, and the way we leverage it is to eliminate unsound and risky decisions and favour those with better outcomes.This is the key to our success and unparalleled excellence in this space. Our people believe in impact, and to make an impact we do not tell people what they already know. Instead we help them cut through the noise in the data, distill down to what matters and encourage them to use the best evidence possible. Our mission is to uplift human outcomes by creating the world's most trustworthy AI engine and products as well as understanding and supporting our clients on this journey of change.