In today’s digital era, the adoption of cutting-edge technologies like AI and Generative AI presents a wealth of benefits across industries and countries. It’s anticipated to significantly boost worker productivity and add trillions to the global economy. But as organizations position themselves to integrate AI, they often encounter challenges in data capture.
Turning to the modality of speech, the Israeli startup aiOla has pioneered the use of NLU (Natural Language Understanding) and ASR (Automatic Speech Recognition) technologies, aiming to equip businesses with the capability to close the data capture gap, streamlining processes and enhancing operational efficiency.
At its core, aiOla’s patented speech AI solution is capable of understanding over 100 languages, filtering through ambient noise, and adapting to various accents and terminologies. Their approach saves time and resources, and empowers a global workforce to operate in their native languages. The startup was founded in 2020 by Guy Ernst and Amir Haramaty, and has attracted over $33 million in VC backing, having raised its Series A in 2022. In our interview with CEO, Haramaty, we explored the vision behind aiOla, exploring the challenges it aims to solve, its impact on the industry, and the future it envisions for speech recognition technology in the business world.
Your journey into the AI and big data realms predates the current explosion of interest. What sparked your interest?
I’ve been immersed in the tech and the big data field for years. Over the past decade, I’ve been deeply involved with AI, well before they became the most used and abused letters in the world. My previous company was the AI platform of choice of one of the leading global management consulting companies where I had the opportunity to engage with some of the leading enterprises in the world.
Through these experiences, I came to understand that while AI and ML are incredible tools, they are just tools and the most important element is data. Separately, only a few, mainly data scientists, were actually benefiting from AI and therefore I wanted to go after the uncaptured and unstructured data that was missing while connecting the promise of AI to the masses.
What convinced you to harness the potential of speech technology?
With the understanding that most data is still untapped and most people in enterprise organizations are not benefiting from AI, I wanted aiOla to address both of these issues. In addition, we realized that speech is the most effective means of communication but due to the limitations and poor accuracy of existing technology, it was not yet incorporated into the data-capturing process. So, at aiOla we went after the speech barrier with the goal of turning words into action while making it accessible and usable for anyone, in any industry, and in any role.
What’s behind aiOla’s technology? How does it differ from the market?
When we understood that most of the data wasn’t part of the game, we realized that speech is the most effective means of communication due to its speed, accuracy, and richness. However, speech isn’t typically considered in the realm of data for AI, especially in business and industrial settings.
ASR faces challenges when dealing with business processes, particularly with jargon and slang. The maximum accuracy achievable with existing ASR technology is subpar and simply not good enough. Speech hasn’t been fully integrated into the AI data game.
aiOla broke through the speech barrier. We incorporated ASR and NLU in any language, any jargon, and across various industries. We’ve developed and registered three patents, and after the soft launch, we attracted top, best-of-breed enterprises from various verticles such as supply chain, financial services, fleet management, food manufacturing, pharmaceuticals, and more.
We realized we are solving something fundamentally different. We’re not aiming to disrupt processes; instead, we’re here to augment and enhance them, making them safer, smarter, more collaborative, and efficient.
The market reaction has been phenomenal. While 2023 was clearly the year of AI, or Generative AI, to be precise, it was fueled by massive hype. In 2024, we expect another banner year, although very different. This year will focus on pragmatic and practical AI that impacts bottom-line ROI and has a massive impact on EBITDA.
We have incredible investors and amazing partners, and the trajectory ahead of us is simply mind-blowing. aiOla brings a unique combination of state-of-the-art technology with simple deployment and onboarding, with a wide horizontal approach. aiOla enables organizations to do more with less and to focus on efficiency, productivity and bottom-line performance.
What are common challenges traditional enterprises face when adopting speech AI?
Our mission is to transform speech in any language, accent and acoustic environment across all workflows in every industry, and to effectively capture data with speech with unprecedented accuracy. We empower organizations at any maturity level to fully leverage the power of AI by utilizing their own data through speech. We turn your words into action, we speak your language.
Until now, the biggest challenge has been accuracy. For instance, we dealt with one of the largest tire companies in the world, and they shared that since 2018, they have been aware that speech is the most efficient way to capture data. They even tried to adopt it themselves but every off-the-shelf ASR platform tested, provided accuracy rates of sub-60%. Hence, despite their commitment, they were unable to implement it. aiOla broke that speech accuracy barrier and demonstrated an ability to do this with accuracy greater than 95% in similar conditions. This is a significant technological breakthrough that shifted the paradigm and built confidence across industries.
How do you see the future applications of aiOla in 5-10 years?
Above and beyond aiOla, the future of AI, and specifically Generative AI, is reshaping everything we do in our lives. aiOla is poised to play a significant role in using speech to drive data-driven actions in any workflow or process within organizations, all without manual intervention. The value is quantifiable, and the future looks very promising. We anticipate playing a small but critical role in every process within organizations, with a measurable impact on ROI.
Advice for fellow startup founders?
Follow the pain trail—don’t develop for the sake of the technology. Begin by identifying customers’ biggest pain points, and focus on creating a must-have product rather than a nice-to-have one. Respect existing processes and approach them with care, adopting a non-threatening approach. Aim to complement existing processes, always prioritizing value and the ability to measure ROI and scale.
