Speechmatics

Everyone's teaching AI to talk. We're teaching it to listen. 👂 Think about the most emotionally intelligent people you know. They're not the ones who talk the most – they're the ones who truly listen. Who catch not just your words, but your tone, your hesitation, your excitement. While the world focuses on making AI sound human, we're working on something more fundamental: making it understand humans better. 👏 At Speechmatics, we're powering the future of voice AI with two game-changing solutions: 🗣Flow: Our end-to-end Conversational AI API for seamless speech interactions  ⏱Real-Time Speech-to-Text: Lightning-fast, accurate speech recognition for any voice stack Discover why we're the best ears in AI: https://lnkd.in/ec_SAQf8 #AI #Speechtech #VoiceAI #ConversationalAI 

Revolutionizing Compliance: Speechmatics & ACA Group join forces 🔊   We're thrilled to welcome ACA Group, the leading GRC advisor in financial services, to our customer family. Our industry-leading speech recognition now powers their RegTech suite, enabling multilingual voice communication monitoring with unprecedented accuracy. 🌐   Starting this week: Join our three-part blog series with ACA's Managing Director Pete Kenny exploring AI transcription in compliance, multilingual surveillance, and the future of voice AI in regulatory oversight.   Keep your eyes on our feed 👀   #AI #RegTech #Finance #Compliance

"AI has gone from black and white to technicolour on steroids" - our CEO Katy Wigdahl shares her insights on the evolving AI scene in PwC's latest series. 📢 While Vegas-style AI dazzle is certainly entertaining, Katy discusses what actually matters: solving real problems for real people. No shiny gimmicks required. At Speechmatics, we're busy making sure every voice is understood - regardless of age, ethnicity, or gender - because we believe AI should work for everyone, not just the easiest to understand. Check out Katy's full interview to discover how we're focusing on the "who" and "how" of speech, not just the "what". 👇 #AI #VoiceAI #ASR

We’re live at #NABShow in Las Vegas and ready for Day 1! Come by and meet the Speechmatics team at our booth in the West Hall – W2317. We’d love to show you what we’ve been working on and chat about how we can support your next project. See you soon! #NAB2025 #Vegas #BroadcastTech

In the series of Conversational AI for Enterprise focus, another problem that needs to be solved: Noise. It's not a new problem. Scientists, sound engineers, electrical engineers, IT teams have been trying to solve this for decades, with the first widely publicised technique being published in 1966 by Dolby. Talking to customers over the last few months this is probably the biggest hurdle (or one of) impeding adoption of Voice Agents. In airports, doctor's clinics, restaurants, conventions, our houses. More than one person's voice renders pretty much all voice AI unusable. At Speechmatics, we tried to solve it a different way. The machine learning way. With Attention. Because that's all you've needed. And that's how we humans do it. Being able to confidently seperate speakers and then identify who to pay attention to is a scalable way of solving this. No need for custom hardware. No need for extra SDKs. Just pure diarisation. Read more about our thought process behind Speaker Lock, on my blog at https://lnkd.in/eB6hGSRJ #gowiththeflow #conversationalai #voiceai #ai

Introducing the missing piece in Voice AI 👀 Voice AI today has a fundamental problem - it simply doesn't know who to listen to. 🎯 In the real world, conversations aren't neat and tidy. They're messy, overlapping, and full of distractions. Yet most voice AI treats all audio input equally, leading to misinterpretations that frustrate users and undermine trust. That's why Speechmatics has developed Speaker Lock technology. Rather than capturing everything, our AI can dynamically identify who it should be listening to and filter out everything else - much like humans naturally do in conversation. 🔊 The applications are incredibly exciting: 📞 Customer service calls remain accurate despite background noise 🍔 Drive-thru orders don't get derailed by backseat chatter 🏥 Medical professionals can dictate notes in busy clinical environments 🚨 Emergency services can capture critical information amid chaos What stands out is our focus on real-world performance rather than controlled lab conditions. It's one thing to achieve high accuracy in perfect testing environments, but quite another to maintain it in the unpredictable settings where voice AI actually needs to work. Find our more with Rohan Sarin’s latest blog: https://lnkd.in/eWbe8MJz #VoiceAI #SpeechRecognition #AI #Innovation

From science fiction to industry standard in under a decade… 🛸 Voice AI has transformed from a futuristic concept to an essential tech stack tool faster than anyone predicted. In one week, we'll be back in Vegas for #NAB2025 to demonstrate just how far this technology has come. Join us in the West Hall (W2317) to experience our AI-powered speech recognition and conversational AI that are setting new accuracy benchmarks in the market. 🗣️ Whether you're struggling with post-production bottlenecks or looking to make your content more accessible, we’ve got solutions worth exploring. Skip the queues and book your demo slot: 👉 https://lnkd.in/dBwC-3pF #NAB #NABShow #VoiceAI #Media 

🥡 Sam Sykes, our Director of Innovation (Head of Tinkering), has been at it again.   He’s built a POC to show how an AI agent could be the future of ordering prawn toast (and yes, it sounds just as fun as it is). Everything you hear in this conversation is powered by Flow—our real-time voice AI agent builder that turns ideas into working prototypes in minutes. Natural, fast, and seriously customizable. Prawn toast today 🍤 The future of voice tomorrow

Want to learn how to achieve hypergrowth and 30x your X/Twitter following in an under an hour? Follow the helpful guide below. 1. Have 1 follower 2. Organise a live space with genuine influencers who have 500k more followers than you do. Sorted Robert Scoble & Irena Cronin thank you for an excellent chat on the future of Conversational AI and kudos on the handling of the very amusing heckler who was "a few Modelo beers down" when he joined. The big takeaway? A truly intelligent voice system starts with genuinely understanding what's being said - context, speakers, emotion and all. If you're not going to listen, there's no point in talking.   Managed to cover how we've evolved from the occasional cable-trip disaster (we don't talk about that time both the kettle and toaster took out ALL the servers...) to building speech tech with 0.6 second latency that can actually tell the difference between speakers.   Favourite bits of the chat: Why speaker diarization (fancy speech word for knowing who's saying what) is critical when you need to focus on one voice in noisy environments How our system won't get confused when kids start shouting for extra fries from the back seat The future of on-device processing (spoiler: we think about 50% of Voice AI is heading that way) My now slightly infamous medical rap demo proving our 2 million word lexicon handles "Bendroflumethiazide" without breaking a sweat   If you want to listen to the full 1hr 15min recording, you can check it out here P.s. if you follow me and we hit the magic 100 followers by next week then I will ensure my first post is a compilation video of the office fails in year 1 at Speechmatics (smoking servers, office cricket, steam roller road melts, kickflip to coffee press and marketing in a box) - @RiquiHerreros https://lnkd.in/eFDmRnD3

Thinking about joining Speechmatics? Here's something for you 👀 We've just launched a careers widget that lets you practice your interview skills before the real thing. It covers everything from those classic "tell me about yourself" questions to the more left-field or technical queries that might come up. Our team is brilliant (and yes, actually lovely) but we know interviews can still be nerve-wracking. This tool helps you arrive prepared and confident. 👏 You can give it a try here: https://lnkd.in/eaDQs4Uu Good luck - we're looking forward to meeting you! #AI #HR #Careers #Hiring