Leonid Joffe, or simply Leo, as most people call him, is a Lead AI Scientist at Brillian. He is the go-to person whenever there’s a complex anomaly detection challenge or a machine learning model that needs to be built to make accurate predictions. Beyond his technical expertise, Leo is also the kind of person with whom it’s almost too easy to drift into a deep conversation about the state of humankind: “I’m fascinated by how technology enables new ways of communicating and interacting — and how that changes the essence of humanity.”, Leo comments.
At Brillian, Leo brings deep research and real-world expertise to the table. He holds a PhD in Software Engineering from University College London and has built a career across companies like Silo AI, NeuralDSP, and Copado before joining Brillian as one of our first employees almost a year ago.
With both academic rigor and industrial practicality, Leo’s talent has been put to use in designing advanced machine learning models for industrial automation, signal processing, and generative AI solutions. His work spans developing anomaly detection systems for manufacturing, optimized software testing pipelines, and building scalable genAI solutions integrating RAG architectures, database optimization, and modular software design.
“As different as my projects may seem, they all share one goal,” Leo explains. “It’s always about creating real value with machine learning or LLMs — not using AI for its own sake. The problems truly benefit of an AI solution.”
Having seen many AI projects, Leo’s advice to business leaders is simple:
“Don’t ask where to add AI — ask what you want to achieve. The term “AI” can be applied to anything from a fancy Excel spreadsheet to ChatGPT. Some of these tools seem like magic, but the better you understand them, the less magical they appear. So, pick an appropriate tool for your task, try to steer clear of the hype bias, and always maintain an understanding of why you’re introducing AI.”
While not at work, Leo explores technology and music in his own terms. To keep up with the rapid evolution of technology, Leo browses Hacker News, and YouTube channels like Two Minute Papers, Fireship, 3blue1brown, and interviews with AI specialists. A lot of cutting-edge research is available in open source, in arXiv and Github repositories. “It’s impossible to keep a detailed track of everything, but following a wide range of resources, even just superficially, helps to build up an intuition of what is going on in the field.”
We live in a fascinating era. “Information overabundance requires both AI and more human critical thinking. The printing press revolutionized how ideas spread, and how communities and institutions are formed around them. On one hand, it crumbled the previous status quo and brought about the 30 years’ war. On the other hand, it was foundational for the era of enlightenment. AI builds on top of the internet, social media, and instant messaging to supercharge the era of massive global communication. It’s significantly disrupting traditional ways of doing things and the world we grew up in. However, it will certainly yield unimaginable advances for humanity. Either way, we are living in exciting, unprecedented* times.”
*oh, what I wouldn’t give for some precedented times!