Olly joined Curinos in 2023 and leads the company's Technology team, focused on helping financial institutions attract, retain and grow more profitable customer relationships; by leveraging AI-based decisioning tools, predictive analytics and Curinos vast proprietary data. Prior to Curinos, Olly held enterprise executive roles leading technology teams in Data & Machine Learning Platforms, Personalization, Marketing Technology and Analytics. Olly was also Founder, CEO of Amplero prior to its acquisition by Novantas; where he helped invent the core technology, established and grew the team and built an investable financial plan and then led the business through spinout and multiple venture capital funding rounds. 

A Journey from Early Machine Learning to Big Data Innovations

I think careers are all about being ready with the right skills when the time comes to catch a “wave”.  I was lucky to be a Physicist in the mid-1990s when some of the early work underpinning what we now call Machine Learning and AI was done and had the privilege of being a Master’s Student of Bayesian Machine Learning pioneer Prof. Sir David MacKay and PhD student of 2024 Nobel Prize winner, Prof. John J. Hopfield.  I would say there have been 4 “waves” of AI, of which I have had the privilege of being around for 3!  I published my first paper on “The Nonnegative Boltzmann Machine” a Generative AI model for natural language, handwriting and human face images in back in 1999 – deemed “interesting, but not very useful” by reviewers at the time!  I went on to focus on novel approaches for solving optimization problems at the heart of training ML and AI models, leading me to invent a type of Quantum Computer, now in use by Google, Lockheed and 3-letter government agencies!

The push to demonstrate interesting applications of these technologies led me towards a career in applying machine learning to the world of burgeoning Big Data, enabling opportunities to fuel businesses in first-of applications to Location Intelligence, Advertising Technology and Marketing Technology and resulting in 43 patents.  Today billions of people experience technologies I have developed on their phones, in their car and even on TikTok!

As the world continues to learn, data is the fuel for intelligent decision-making, whether by humans or machines – Curinos is at the heart of banking data in North America; leveraging it into our software solutions that deliver on our mission to leverage proprietary data, decision tools and A.I. to help our clients optimize their go-to-market decisions.  A natural fit for my background and expertise.

Ensuring Compliance and Security in AI

Financial institutions rightly operate in a highly regulated environment, in which managing data privacy and security and addressing and controlling for bias and disparate impact are critical considerations.  Such constraints crystallize key considerations for almost all AI applications, the relationship between human and the AI and the need to provide “control surfaces” to technologies you build with AI that provide hard guarantees of compliance, safety and customer or user experience – not easy with a technology that is by definition probabilistic, rather than deterministic.  In our products at Curinos we have made the setting of constraints, human review and approval workflows and rigorous audibility and observability to prove the constraints are met, a seamlessly integrated element to the bank customer experience.  Our AI capabilities help people make good decisions from a bounded universe of acceptable ones, having a dramatic impact on the outcome of those decisions for both the bank and their customers.

The Role of Patents in Advancing Quantum Optimization and AI

It’s simple – patents help protect your ability to do business with your investments in innovation – for me as a technologist, they replace academic publication as a validated record of serial innovation.  When we believe we have clarity on how to put into practice an idea that will give us a business advantage for years to come, we invest in protecting it.  For early-stage products or lines of business this protection adds value in advance of significant commercial traction from the innovation – the intellectual property.

The Future of Financial Services

In financial services, as in many consumer-facing industries, I think we are seeing the first wave of Generative AI-enabled capabilities pass through domains like customer service where Retrieval Augmented Generation patterns can really help enable conversational interaction with factual information and data, improving on previous Online Chat and IVR solutions, benefiting the customer experience and reducing cost.  In the next wave, I expect the convergence of language intelligence with analytic intelligence, powering bank functions and the customer experience and bringing a bank’s business objectives and the customer experience together in how bank products are structured, priced, delivered and proposed to customers, whether that customer is a consumer, small business or enterprise.  There are some natural starting points here, in pricing optimization, customer-offer decisioning and driving branch performance, where the combination of digital and physical banking experiences for clients come together.  Financial Services is also a heavily analytical industry in general and I expect the user experience bar and accessibility for non-technical personas of both application-specific and general-purpose analytics to advance dramatically within the next 12-18 months.

Advice for Aspiring Technologists and Entrepreneurs in AI and Machine Learning

Sweat the details! The devil is in the details of the 20% from the 80/20 of going from a cute demo to having built a truly productized AI/ML capability; operations, optimization, observability, safety, governance and guardrails all come into play with a truly useful and scalable operationalized ML/AI solution.