Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur.
MiFID II and FCA call recording: compliance for voice transcription in finance
TL;DR: Financial firms operating under MiFID II and FCA jurisdictions must maintain searchable, high-accuracy records of all client-related voice communications, including remote and mobile calls under FCA Market Watch 66. Engineering and product teams building for these obligations typically implement dedicated cloud infrastructure with defined data residency, controls that prevent customer audio from being used to retrain models on Growth and Enterprise plans, and transcription accurate enough that the resulting records hold up under regulatory review. Transcription and speaker attribution errors are not product quality issues in this context. They are audit trail failures, and regulators treat them as such.
Why speech-to-text accuracy matters upstream of your LLM
TL;DR: Downstream LLM performance is ceiling-bounded by upstream transcription accuracy. A transcript with a meaningful error rate doesn't produce proportionally degraded summaries or CRM entries. It produces outputs where hallucinated names, inverted logic, and misattributed speaker turns compound silently into every downstream system that reads them. Prompt engineering cannot recover information the STT layer never captured. Gravite cut call quality review time by 93%, from 15 minutes to 1 minute per call, once transcript accuracy was high enough to trust the output without manual verification.
TL;DR: Telephony audio is constrained to 8kHz narrowband frequencies, stripping away the high-frequency spectral energy that generic 16kHz STT models require. Standard upsampling cannot recover phonemes that were never captured at the source, leading to transcription errors that compound silently into downstream NLU failures. Solaria-3 ranks #1 on Switchboard, the most challenging conversational telephony dataset, ahead of AssemblyAI, ElevenLabs, Deepgram, Mistral, and Speechmatics, ensuring your downstream LLM pipelines receive clean, structured data from real-world noisy call audio.
Recall and Gladia join forces to power online meetings transcription
Published on Oct 19, 2023
Today, we are thrilled to announce a partnership aimed at empowering businesses and developers worldwide to fully leverage data from online meetings.
Recall, a pioneering developer tooling, API, and infrastructure provider best known for plug-and-play meeting bots, has teamed up with Gladia to provide real-time code-switching and accurate transcription to over 100 clients worldwide.
Recall: Capturing the essence of meetings
As the world grappled with the COVID-19 pandemic, the demand for video conferencing solutions skyrocketed, multiplying the number of Zroom calls alone by an astonishing 100-fold.
Founded in February 2022, Recall’s mission was to provide companies worldwide with the best possible infrastructure powered by LLMs to extract valuable data from virtual meetings.
Recall allows developers to build products on top of meeting data captured from key platforms like Zoom, Google Meet, and others. They offer a comprehensive API that enables video and audio recordings, transcriptions, and metadata extractions (participant names, timestamps, etc.)
While it takes at least six months on average to develop meeting bots in-house, with Recall, companies can seamlessly integrate these functionalities in a matter of days.
Owing to its versatility and ease of use, Recall has exhibited spectacular growth and now caters to a wide range of enterprise clients across various industries and use cases, including sales enablement tools, note-taking solutions, productivity-enhancing applications, and more.
Gladia x Recall: Advancing meeting data transcription
Transcription is a critical component of video recording and conferencing tools provided by Recall.
At Gladia, we built an enterprise version of OpenAI’s Whisper ASR in the form of an API, distinguished by exceptional accuracy and speed, extended language support, and a variety of additional features.
Virtual meeting and note-taking have been among the most important use cases for Gladia, making our API a perfect candidate to address the challenges of virtual meeting transcription.
With Gladia's API integration, Recall's clients can now directly enjoy the benefits of instantaneous and accurate meeting transcription, including extended language support, speaker diarization, and word-level timestamps.
We’re grateful for the trust and thrilled to partner with a company like Recall, whose ambition to help companies improve the way they work by leveraging data from meetings aligns perfectly with Gladia’s vision and objectives.
For a more detailed practical tutorial on using Gladia API with Recall’s meeting bots, head to the tutorial on Recall’s website.
About Gladia
At Gladia, we built an optimized version of Whisper in the form of an API, adapted to real-life professional use cases and distinguished by exceptional accuracy, speed, extended multilingual capabilities and state-of-the-art features.
Contact us
Your request has been registered
A problem occurred while submitting the form.
Read more
Speech-To-Text
MiFID II and FCA call recording: compliance for voice transcription in finance
Speech-To-Text
Why speech-to-text accuracy matters upstream of your LLM
