Jean-Louis Quéguiner is the Founder and CEO of Gladia. He beforehand served as Group Vice President of Knowledge, AI, and Quantum Computing at OVHcloud, one in all Europe’s main cloud suppliers. He holds a Grasp’s Diploma in Symbolic AI from the College of Québec in Canada and Arts et Métiers ParisTech in Paris. Over the course of his profession, he has held important positions throughout varied industries, together with monetary knowledge analytics, machine studying functions for real-time digital promoting, and the event of speech AI APIs.
Gladia offers superior audio transcription and real-time AI options for seamless integration into merchandise throughout industries, languages, and know-how stacks. By optimizing state-of-the-art ASR and generative AI fashions, it ensures correct, lag-free speech and language processing. Gladia’s platform additionally allows real-time extraction of insights and metadata from calls and conferences, supporting key enterprise use instances reminiscent of gross sales help and automatic buyer help.
What impressed you to sort out the challenges in speech-to-text (STT) know-how, and what gaps did you see out there?
Once I based Gladia, the preliminary objective was broad—an AI firm that might make advanced know-how accessible. However as we delved deeper, it grew to become clear that voice know-how was probably the most damaged and but most important space to deal with.
Voice is central to our each day lives, and most of our communication occurs by way of speech. But, the instruments out there for builders to work with voice knowledge have been insufficient by way of velocity, accuracy, and value—particularly throughout languages.
I needed to repair that, to unpack the complexity of voice know-how and repackage it into one thing easy, environment friendly, highly effective and accessible. Builders shouldn’t have to fret concerning the intricacies of AI fashions or the nuances of context size in speech recognition. My objective was to create an enterprise-grade speech-to-text API that labored seamlessly, whatever the underlying mannequin or know-how—a real plug-and-play answer.
What are among the distinctive challenges you encountered whereas constructing a transcription answer for enterprise use?
In terms of speech recognition, velocity and accuracy—the 2 key efficiency indicators on this discipline—are inversely proportional by design. Because of this enhancing one will compromise the opposite, a minimum of to some extent. The price issue, to a giant extent, outcomes from the supplier’s selection between velocity and high quality.
When constructing Gladia, our objective was to search out the right steadiness between these two components, all whereas guaranteeing the know-how stays out there to startups and SMEs. Within the course of we additionally realized that the foundational ASR fashions like OpenAI’s Whisper, which we labored with extensively, are biased, skewering closely in direction of English resulting from their coaching knowledge, which leaves a number of languages under-represented.
So, along with fixing the speed-accuracy tradeoff, it was vital to us— as a European, multilingual group—to optimize and fine-tune our core fashions to construct a really international API that helps companies function throughout languages.
How does Gladia differentiate itself within the crowded AI transcription market? What makes your Whisper-Zero ASR distinctive?
Our new real-time engine (Gladia Actual Time) achieves an industry-leading 300 ms latency. Along with that, it’s capable of extract insights from a name or assembly with the so-called “audio intelligence” add-ons or options, like named entity recognition (NER) or sentiment evaluation.
To our data, only a few rivals are capable of present each transcription and insights at such excessive latency (lower than 1s end-to-end) – and do all of that precisely in languages aside from English. Our languages help extends to over 100 languages right now.
We additionally put a particular emphasis on making the product really stack agnostic. Our API is suitable with all present tech stacks and telephony protocols, together with SIP, VoIP, FreeSwitch and Asterisk. Telephony protocols are particularly advanced to combine with, so we consider this product side can convey super worth to the market.
Hallucinations in AI fashions are a big concern, particularly in real-time transcription. Are you able to clarify what hallucinations are within the context of STT and the way Gladia addresses this downside?
Hallucination normally happens when the mannequin lacks data or doesn’t have sufficient context on the subject. Though fashions can produce outputs tailor-made to a request, they will solely reference data that existed on the time of their coaching, and that might not be up-to-date. The mannequin will create coherent responses by filling in gaps with data that sounds believable however is inaccurate.
Whereas hallucinations grew to become identified within the context of LLMs first, they happen with speech recognition fashions— like Whisper ASR, a number one mannequin within the discipline developed by OpenAI – as nicely. Whisper’s hallucinations are like these of LLMs resulting from an identical structure, so it’s an issue that issues generative fashions, which might be capable of predict the phrases that comply with based mostly on the general context. In a method, they ‘invent’ the output. This strategy might be contrasted with extra conventional, acoustic-based ASR architectures that match the enter sound to output in a extra mechanical method
In consequence, chances are you’ll discover phrases in a transcript that weren’t really mentioned, which is clearly problematic, particularly in fields like drugs, the place a mistake of this sort can have grave penalties.
There are a number of strategies to handle and detect hallucinations. One widespread strategy is to make use of a retrieval-augmented technology (RAG) system, which mixes the mannequin’s generative capabilities with a retrieval mechanism to cross-check info. One other methodology includes using a “chain of thought” strategy, the place the mannequin is guided by way of a collection of predefined steps or checkpoints to make sure that it stays on a logical path.
One other technique for detecting hallucinations includes utilizing programs that assess the truthfulness of the mannequin’s output throughout coaching. There are benchmarks particularly designed to judge hallucinations, which contain evaluating completely different candidate responses generated by the mannequin and figuring out which one is most correct.
We at Gladia have experimented with a mix of methods when constructing Whisper-Zero, our proprietary ASR that removes nearly all hallucinations. It’s confirmed glorious leads to asynchronous transcription, and we’re at the moment optimizing it for real-time to attain the identical 99.9% data constancy.
STT know-how should deal with a variety of complexities like accents, noise, and multi-language conversations. How does Gladia strategy these challenges to make sure excessive accuracy?
Language detection in ASR is an especially advanced activity. Every speaker has a novel vocal signature, which we name options. By analyzing the vocal spectrum, machine studying algorithms can carry out classifications, utilizing the Mel Frequency Cepstral Coefficients (MFCC) to extract the primary frequency traits.
MFCC is a technique impressed by human auditory notion. It’s a part of the “psychoacoustic” discipline, specializing in how we understand sound. It emphasizes decrease frequencies and makes use of methods like normalized Fourier decomposition to transform audio right into a frequency spectrum.
Nonetheless, this strategy has a limitation: it is based mostly purely on acoustics. So, for those who converse English with a robust accent, the system might not perceive the content material however as a substitute choose based mostly in your prosody (rhythm, stress, intonation).
That is the place Gladia’s progressive answer is available in. We have developed a hybrid strategy that mixes psycho-acoustic options with content material understanding for dynamic language detection.
Our system would not simply hearken to the way you converse, but in addition understands what you are saying. This twin strategy permits for environment friendly code-switching and would not let sturdy accents get misrepresented/misunderstood.
Code-switching—which is amongst our key differentiators—is a very vital characteristic in dealing with multilingual conversations. Audio system might swap between languages mid-conversation (and even mid-sentence), and the power of the mannequin to transcribe precisely on the fly regardless of the swap is important.
Gladia API is exclusive in its means to deal with code-switching with this many language pairs with a excessive stage of accuracy and performs nicely even in noisy environments, identified to scale back the standard of transcription.
Actual-time transcription requires ultra-low latency. How does your API obtain lower than 300 milliseconds latency whereas sustaining accuracy?
Retaining latency underneath 300 milliseconds whereas sustaining excessive accuracy requires a multifaceted strategy that blends {hardware} experience, algorithm optimization, and architectural design.
Actual-time AI isn’t like conventional computing—it’s tightly linked to the facility and effectivity of GPGPUs. I’ve been working on this area for almost a decade, main the AI division at OVHCloud (the largest cloud supplier within the EU), and discovered firsthand that it’s all the time about discovering the appropriate steadiness: how a lot {hardware} energy you want, how a lot it prices, and the way you tailor the algorithms to work seamlessly with that {hardware}.
Efficiency in actual time AI comes from successfully aligning our algorithms with the capabilities of the {hardware}, guaranteeing each operation maximizes throughput whereas minimizing delays.
However it’s not simply the AI and {hardware}. The system’s structure performs a giant position too, particularly the community, which may actually impression latency. Our CTO, who has deep experience in low-latency community design from his time at Sigfox (an IoT pioneer), has optimized our community setup to shave off useful milliseconds.
So, it’s actually a mixture of all these components—sensible {hardware} decisions, optimized algorithms, and community design—that lets us constantly obtain sub-300ms latency with out compromising on accuracy.
Gladia goes past transcription with options like speaker diarization, sentiment evaluation, and time-stamped transcripts. What are some progressive functions you’ve seen your shoppers develop utilizing these instruments?
ASR unlocks a variety of functions to platforms throughout verticals, and it’s been superb to see what number of really pioneering corporations have emerged within the final two years, leveraging LLMs and our API to construct cutting-edge, aggressive merchandise. Listed below are some examples:
Sensible note-taking: Many purchasers are constructing instruments for professionals who have to shortly seize and arrange data from work conferences, scholar lectures, or medical consultations. With speaker diarization, our API can determine who mentioned what, making it simple to comply with conversations and assign motion objects. Mixed with time-stamped transcripts, customers can soar straight to particular moments in a recording, saving time and guaranteeing nothing will get misplaced in translation.Gross sales enablement: Within the gross sales world, understanding buyer sentiment is the whole lot. Groups are utilizing our sentiment evaluation characteristic to achieve real-time insights into how prospects reply throughout calls or demos. Plus, time-stamped transcripts assist groups revisit key components of a dialog to refine their pitch or deal with shopper issues extra successfully. For this use case particularly, NER can also be key to figuring out names, firm particulars, and different data that may be extracted from gross sales calls to feed the CRM robotically.Name middle help: Corporations within the contract middle area are utilizing our API to supply dwell help to brokers, in addition to flagging buyer sentiment throughout calls. Speaker diarization ensures that issues being mentioned are assigned to the appropriate particular person, whereas time-stamped transcripts allow supervisors to evaluation important moments or compliance points shortly. This not solely improves the shopper expertise – with higher on-call decision fee and high quality monitoring – but in addition boosts agent productiveness and satisfaction.
Are you able to talk about the position of customized vocabularies and entity recognition in enhancing transcription reliability for enterprise customers?
Many industries depend on specialised terminology, model names, and distinctive language nuances. Customized vocabulary integration permits the STT answer to adapt to those particular wants, which is essential for capturing contextual nuances and delivering output that precisely displays what you are promoting wants. As an illustration, it permits you to create a listing of domain-specific phrases, reminiscent of model names, in a selected language.
Why it’s helpful: Adapting the transcription to the precise vertical permits you to decrease errors in transcripts, attaining a greater consumer expertise. This characteristic is particularly important in fields like drugs or finance.
Named entity recognition (NER) extracts and identifies key data from unstructured audio knowledge, reminiscent of names of individuals, organizations, places, and extra. A standard problem with unstructured knowledge is that this important data isn’t readily accessible—it is buried inside the transcript.
To unravel this, Gladia developed a structured Key Knowledge Extraction (KDE) strategy. By leveraging the generative capabilities of its Whisper-based structure—just like LLMs—Gladia’s KDE captures context to determine and extract related data instantly.
This course of might be additional enhanced with options like customized vocabulary and NER, permitting companies to populate CRMs with key knowledge shortly and effectively.
In your opinion, how is real-time transcription reworking industries reminiscent of buyer help, gross sales, and content material creation?
Actual-time transcription is reshaping these industries in profound methods, driving unimaginable productiveness beneficial properties, coupled with tangible enterprise advantages.
First, real-time transcription is a game-changer for help groups. Actual-time help is vital to enhancing the decision fee because of quicker responses, smarter brokers, and higher outcomes (by way of NSF, deal with occasions, and so forth). As ASR programs get higher and higher at dealing with non-English languages and performing real-time translation, contact facilities can obtain a really international CX at decrease margins.
In gross sales, velocity and spot-on insights are the whole lot. Equally to what occurs with name brokers, real-time transcription is what equips them with the appropriate insights on the proper time, enabling them to deal with what issues probably the most in closing offers.
For creators, real-time transcription is probably much less related right now, however nonetheless filled with potential, particularly in relation to dwell captioning and translation throughout media occasions. Most of our present media clients nonetheless want asynchronous transcription, as velocity is much less important there, whereas accuracy is vital for functions like time-stamped video enhancing and subtitle technology.
Actual-time AI transcription appears to be a rising pattern. The place do you see this know-how heading within the subsequent 5-10 years?
I really feel like this phenomenon, which we now name real-time AI, goes to be all over the place. Basically, what we actually discuss with right here is the seamless means of machines to work together with folks, the way in which we people already work together with each other.
And for those who take a look at any Hollywood film (like Her) set sooner or later, you’ll by no means see anybody there interacting with clever programs through a keyboard. For me, that serves as the final word proof that within the collective creativeness of humanity, voice will all the time be the first method we work together with the world round us.
Voice, as the primary vector to mixture and share human data, has been a part of human tradition and historical past for for much longer than writing. Then, writing took over as a result of it enabled us to protect our data extra successfully than counting on the group elders to be the guardians of our tales and knowledge.
GenAI programs, able to understanding speech, producing responses, and storing our interactions, introduced one thing fully new to the area. It’s the very best of each phrases and the very best of humanity actually. It offers us this distinctive energy and power of voice communication with the advantage of reminiscence, which beforehand solely written media might safe for us. That is why I consider it’s going to be all over the place – it is our final collective dream.
Thanks for the good interview, readers who want to be taught extra ought to go to Gladia.
