Gladia, an AI-powered transcription and audio intelligence provider based in Paris, France, has successfully secured $16 million in funding. The company plans to leverage this investment to enhance its audio infrastructure, beginning with an innovative real-time audio transcription and analytics engine. This technology aims to empower voice-first platforms with advanced AI capabilities to deliver superior user experiences across various regions.
In an interview, CEO Jean-Louis Quéguiner discussed the motivation behind founding Gladia, highlighting his own challenges with existing transcription services while speaking English with a distinct French accent. "As you can hear from my beautiful French accent," Quéguiner explained, "I was extremely frustrated with the inaccuracies consistently produced by these services."
Quéguiner demonstrated Gladia’s real-time transcribing capabilities, showcasing its impressive accuracy even with his heavy accent. During the demo, the system seamlessly switched between languages, illustrating its multilingual proficiency.
The funding round was led by XAnge, with participation from notable investors such as Illuminate Financial, XTX Ventures, Athletico Ventures, Gaingels, Mana Ventures, Motier Ventures, Roosh Ventures, and Soma Capital. This recent round brings Gladia's total funding to $20.3 million, following earlier seed investments from firms like New Wave, Sequoia Capital, Cocoa, and GFC. The company has also been selected for participation in the AWS Generative AI Accelerator program.
According to Alexis du Peloux, a partner at XAnge, "Gladia embodies the qualities we champion: a bold, global tech team at the forefront of AI innovation, with a proven business model capable of unlocking new opportunities across industries." He commended Quéguiner and his team for their effective execution in a fast-paced AI landscape.
Gladia sets itself apart by prioritizing the development of a truly multilingual product, thus addressing the inherent biases found in most speech recognition models, which are frequently trained primarily on English data. The new engine allows for real-time transcription in over 100 languages, supports accents effectively, and can adapt to language switching on the fly.
What makes Gladia particularly innovative is its ability to extract actionable insights from calls, such as sentiment analysis, key information, and summaries, all generated in less than a second.
Building a real-time, low-latency, multilingual engine is a complex endeavor that demands extensive language understanding and data processing skills. However, Gladia’s product mitigates these challenges. Its speech-to-text engine achieves industry-leading latency of under 300 milliseconds while maintaining high accuracy, regardless of language or technical environment.
"Many companies are investing significant time and resources to integrate various AI functions into their existing platforms," remarked Jonathan Soto, CTO of Gladia. "Our single API is compatible with all tech stacks, including SIP, VoIP, FreeSwitch, and Asterisk, facilitating easy integration of real-time transcription into clients' AI platforms."
Since the launch of its initial async transcription and audio intelligence API in June 2023, Gladia has gained significant traction in the enterprise sector, particularly among meeting recording and note-taking solutions. The API has attracted over 600 global customers, including companies like Attention, Circleback, Method Financial, Recall, Sana, and VEED.IO, and boasts more than 70,000 users.
"Gladia’s technology empowers businesses in sectors requiring advanced real-time transcription, enabling them to transition from manual post-call processing to proactive, low-latency workflows," Quéguiner stated. "Our solutions support automated CRM enrichment and provide real-time assistance for support agents, enhancing operational efficiency without necessitating extensive in-house AI expertise."
With the new funding, Gladia aims to accelerate its R&D efforts, introducing a comprehensive AI toolkit for audio, alongside additional à la carte models such as large language models (LLMs) and retrieval-augmented generation (RAG). The company is currently piloting an agent-assist solution within the contact-center-as-a-service (CCaaS) sector and plans to expand its talent pool in preparation for international growth.
Quéguiner highlighted Gladia's unique capability to perform "code switching," allowing users to initiate transcription in one language and transition seamlessly to another during a call. He emphasized the importance of real-time processing and the ability to integrate custom vocabularies, asserting, "Our model offers real-time insights that set us apart from competitors."
In the realm of market opportunities, Queguiner noted Gladia’s focus on meeting recorders and its collaboration with contact centers, which enhance call completion speeds by as much as 30% due to improved accuracy. Gladia will operate on a straightforward pricing structure, typically charging a flat rate based on hourly usage.
Overall, Gladia stands at the intersection of AI innovation and real-time transcription, poised to transform how businesses engage with audio intelligence in diverse settings.
Ranking
