Denis Ignatovich, Co-founder and Co-CEO of Imandra, has over a decade of expertise in buying and selling, threat administration, quantitative modeling, and sophisticated buying and selling system design. Earlier than founding Imandra, he led the central threat buying and selling desk at Deutsche Financial institution London, the place he acknowledged the crucial function AI can play within the monetary sector. His insights throughout this time helped form Imandra’s suite of monetary merchandise. Denis’ contributions to computational logic for monetary buying and selling platforms embrace a number of patents. He holds an MSc in Finance from the London College of Economics and levels in Laptop Science and Finance from UT Austin.
Imandra is an AI-powered reasoning engine that makes use of neurosymbolic AI to automate the verification and optimization of advanced algorithms, significantly in monetary buying and selling and software program methods. By combining symbolic reasoning with machine studying, it enhances security, compliance, and effectivity, serving to establishments cut back threat and enhance transparency in AI-driven decision-making.
What impressed you and Dr. Grant Passmore to co-found Imandra, and the way did your backgrounds affect the imaginative and prescient for the corporate?
After faculty I went into quantitative buying and selling and ended up in London. Grant did his PhD in Edinburgh after which moved to Cambridge to work on functions of automated logical reasoning for evaluation of security of autopilot methods (advanced algorithms which contain nonlinear computation). In my work, I additionally handled advanced algorithms with numerous nonlinear computation and we realized that there’s a deep connection between these two fields. The best way that finance was creating such algorithms was actually problematic (as highlighted by many information tales coping with “algo glitches”), so we got down to change that by empowering engineers in finance with automated logical instruments to carry rigorous scientific methods to the software program design and improvement. Nevertheless, what we ended up creating is industry-agnostic.
Are you able to clarify what neurosymbolic AI is and the way it differs from conventional AI approaches?
The sphere of AI has (very roughly!) two areas: statistical (which incorporates LLMs) and symbolic (aka automated reasoning). Statistical AI is unbelievable at figuring out patterns and doing translation utilizing info it realized from the information it was skilled on. However, it’s unhealthy at logical reasoning. The symbolic AI is nearly the precise reverse – it forces you to be very exact (mathematically) with what you’re making an attempt to do, however it may well use logic to purpose in a method that’s (1) logically constant and (2) doesn’t require information for coaching. The methods combining these two areas of AI are referred to as “neurosymbolic”. One well-known utility of this method is the AlphaFold mission from DeepMind which just lately received the Nobel prize.
What do you assume units Imandra aside in main the neurosymbolic AI revolution?
There are various unbelievable symbolic reasoners on the market (most in academia) that concentrate on particular niches (e.g. protein folding), however Imandra empowers builders to investigate algorithms with unprecedented automation which has a lot better functions and better goal audiences than these instruments.
How does Imandra’s automated reasoning eradicate frequent AI challenges, comparable to hallucinations, and enhance belief in AI methods?
With our method, LLMs are used to translate people’ requests into formal logic which is then analyzed by the reasoning engine with full logical audit path. Whereas translation errors might happen when utilizing the LLM, the consumer is supplied with a logical clarification of how the inputs have been translated and the logical audits could also be verified by third celebration open supply software program. Our final aim is to carry actionable transparency, the place the AI methods can clarify their reasoning in a method that’s independently logically verifiable.
Imandra is utilized by Goldman Sachs and DARPA, amongst others. Are you able to share a real-world instance of how your know-how solved a fancy drawback?
An ideal public instance of the actual world impression of Imandra is highlighted in our UBS Way forward for Finance competitors 1st place win (the main points with Imandra code is on our web site). Whereas making a case examine for UBS that encoded a regulatory doc that they submitted to the SEC, Imandra recognized a basic and delicate flaw within the algorithm description. The flaw stemmed from delicate logical circumstances that need to be met to rank orders inside an order ebook – one thing that might be not possible for people to detect “by hand”. The financial institution awarded us 1st place (out of greater than 620 corporations globally).
How has your expertise at Deutsche Financial institution formed Imandra’s functions in monetary methods, and what’s essentially the most impactful use case you have seen to date?
At Deutsche Financial institution we handled loads of very advanced code that made automated buying and selling choices primarily based on varied ML inputs, threat indicators, and many others. As any financial institution, we additionally needed to abide by quite a few rules. What Grant and I noticed was that this, on a mathematical degree, was similar to the analysis he was doing for autopilot security.
Past finance, which industries do you see as having the best potential to learn from neurosymbolic AI?
We’ve seen AlphaFold get the Nobel prize, so let’s undoubtedly rely that one… Finally, most functions of AI will tremendously profit by use of symbolic strategies, however particularly, we’re engaged on the next brokers that we are going to launch quickly: code evaluation (translating supply code into mathematical fashions), creating rigorous fashions from English-prose specs, reasoning about SysML fashions (language used to explain methods in safety-critical industries) and enterprise course of automation.
Imandra’s area decomposition is a novel function. Are you able to clarify the way it works and its significance in fixing advanced issues?
A query that each engineer thinks about when writing software program is “what the sting circumstances?”. When their job is QA and they should write unit check circumstances or they’re writing code and fascinated by whether or not they’ve accurately carried out the necessities. Imandra brings scientific rigor to reply this query – it treats the code as a mathematical mannequin and symbolically analyzes all of its edge circumstances (whereas producing a proof in regards to the completeness of protection). This function relies on a mathematical approach referred to as ‘Cylindrical Algebraic Decomposition’, which we’ve “lifted” to algorithms at massive. It has saved numerous hours for our prospects in finance and uncovered crucial errors. Now we’re bringing this function to engineers in all places.
How does Imandra combine with massive language fashions, and what new capabilities does this unlock for generative AI?
LLMs and Imandra work collectively to formalize human enter (whether or not it’s supply code, English prose, and many others), purpose about it after which return the output in a method that’s simple to grasp. We use agentic frameworks (e.g. Langgraph) to orchestrate this work and ship the expertise as an agent that our prospects can use instantly, or combine into their functions or brokers. This symbiotic workflow addresses most of the challenges of utilizing LLM-only AI instruments and extends their utility past beforehand seen coaching information.
What’s your long-term imaginative and prescient for Imandra, and the way do you see it reworking AI functions throughout industries?
We predict neurosymbolic methods would be the basis that paves the best way for us to comprehend the promise of AI. Symbolic methods are the lacking ingredient for many of the industrial functions of AI and we’re excited to be on the forefront of this subsequent chapter of AI.
Thanks for the good interview, readers who want to study extra ought to go to Imandra.