Top Builders

Explore the top contributors showcasing the highest number of app submissions within our community.

Speechmatics

Founded in 2006 by Dr. Tony Robinson, a Cambridge University speech recognition pioneer, Speechmatics builds infrastructure to understand every voice globally. The company's core mission is inclusive, multilingual speech AI, covering transcription, real-time voice agents, and on-device deployment. Speechmatics serves enterprise clients across media, healthcare, financial services, and contact centers.

General
CompanySpeechmatics
Founded2006 by Dr. Tony Robinson
CEOKaty Wigdahl
HeadquartersCambridge, United Kingdom
Websitespeechmatics.com
Documentationdocs.speechmatics.com
GitHubgithub.com/speechmatics
TypeSpeech AI / B2B SaaS

Core Products

Speechmatics API (Speech-to-Text)

The Speechmatics API provides batch and real-time transcription across 55+ languages, powered by the Ursa 2 model released in October 2024. It supports speaker diarization, custom dictionaries, automatic translation, and Voice Intelligence add-ons (summarization, sentiment analysis, entity recognition) with no retraining required.

Speechmatics Flow

Flow is a voice agent API that combines Speechmatics' speech-to-text with an LLM and text-to-speech into a single real-time pipeline. Developers connect through a single API call to build conversational AI agents with smart turn detection, interruption handling, and function calling support.

Developer Resources

Speechmatics provides SDKs for Python, JavaScript/TypeScript, and .NET, along with a developer portal and free tier to get started without a credit card.

  • Documentation — official API reference, quickstarts, and integration guides
  • GitHub — open-source SDKs and client libraries
  • Developer Portal — API key management and usage dashboard
  • Pricing — free tier and pay-as-you-go rates

Key Features

Multilingual transcription across 55+ languages Speechmatics' Ursa 2 model leads accuracy benchmarks in 62% of supported languages on the FLEURS dataset, with 7.88% WER on Kincaid46 for English, surpassing human-level accuracy on that benchmark.

Flexible deployment Speechmatics runs on private SaaS cloud, on-premises, on-device, and via Docker or Kubernetes, making it suitable for data-sensitive industries like healthcare and finance.

Voice Intelligence add-ons Summarization, sentiment analysis, topic detection, chapter generation, and entity recognition layer on top of transcription without requiring additional integration work.

Use Cases

Contact center automation Real-time transcription and sentiment analysis during calls, combined with Flow for automated voice agent handling of common queries.

Clinical transcription Speechmatics' Medical Model (launched 2024) targets 93% real-time accuracy and 96% medical keyword recall for English, German, Danish, and Norwegian.

Media and broadcast Batch transcription of audio and video files for subtitling, archiving, and content search across multiple languages.

Edit on GitHub

speechmatics AI Technologies Hackathon projects

Discover innovative solutions crafted with speechmatics AI Technologies, developed by our community members during our engaging hackathons.

MidContext Live Translation Agent

MidContext Live Translation Agent

MidContext Live Translation Agent solves a major challenge for companies operating across multilingual markets: customer support becomes slower, more expensive and less personal as customers and agents do not speak the same jargon. Beyond language, each generation have its unique way of talking and AI enables hyper customisation capabilities. We identified low scalable workflows, high wait times, low resolution quality and inconsistent customer experience as key pain points for companies, especially for companies scaling across Europe with different languages, accents and local expectations, and low maturity with their internal knowledge bases. Scalable globally, and also interesting to mayor incumbents that can not afford losses in their reputation. Our solution is a real-time voice translation layer between customer care agents and customers. The system captures voice input, converts speech through ASR, routes the conversation through a customer support layer, and generates natural voice responses using TTS. It does more than translate words: it preserves context, intent, tone and company jargon, while connecting to local knowledge bases and support workflows. It works today, right away in the company as it is, and help build its future enriching their local customer service knowledge base. The target users are multinational companies, customer operations teams, CCaaS providers and enterprises that need scalable multilingual support without losing the human connection. MidContext uses a glocal strategy: one global architecture, adapted to local languages, customer behaviors, policies and knowledge bases. A human-in-the-loop quality model keeps agents responsible for sensitive cases, approvals and escalations, reducing technological complexity while improving trust, resolution quality and customer satisfaction.

CartMate AI Shopping Assistant

CartMate AI Shopping Assistant

CartMate is an AI-powered conversational commerce assistant designed to make online shopping easier, faster, and more accessible. Many users struggle with traditional text-based search because they may not know the right keywords, product category, brand name, or technical terms. CartMate solves this problem by allowing users to simply speak what they want in natural language. The system understands the user’s intent, asks smart follow-up questions, and recommends suitable products from the catalogue. CartMate also includes visual product matching. Users can show a physical item or upload an image, and the AI analyzes its category, color, material, style, and visible branding. Based on this analysis, CartMate finds visually similar products and displays ranked recommendations. This is especially useful for elderly users, low-literacy users, non-technical shoppers, and customers who can recognize a product visually but cannot describe it accurately. The project uses real-time speech-to-text, conversational AI reasoning, text-to-speech output, and AI vision to create a smooth shopping experience. For businesses, CartMate improves product discovery, reduces search frustration, increases customer engagement, and can help improve conversion rates. It can be integrated as an intelligent shopping layer for e-commerce platforms, marketplaces, and online stores.

real time translator/transcriber

real time translator/transcriber

RealTime Speech Transcription & Translation This project implements a real-time speech transcription and translation pipeline, exposed through a lightweight web service built with FastAPI and Uvicorn. The backend provides a set of REST endpoints for lifecycle management (load, unload, reset) and a WebSocket endpoint for continuous audio streaming and live transcription. Audio is captured directly from the browser microphone using the WebAudio API, downsampled to 16kHz mono PCM, and streamed in chunks to the backend over WebSocket. The server processes each chunk through a transcription engine and a translation pipeline, returning results in real time. The architecture supports two interchangeable transcription backends: a cloud-based path using the Speechmatics Real-Time API, and a fully local path using faster-whisper (large-v3) running on a dedicated NVIDIA RTX 3090 GPU. Translation is handled by facebook/nllb-200-1.3B, supporting Arabic, Italian, English, and French as both source and target languages. The translator is designed around a pluggable interface, making it straightforward to add new language pairs or swap the underlying model. A minimal Vue 3 single-page application provides the client interface, with live panels showing the accumulated transcription, the current in-progress line, and the raw original text before translation. Demo running in local mode only.

Synapse Corp AI

Synapse Corp AI

Synapse AI is an enterprise-grade multi-agent workflow automation platform designed to simulate how real organizations operate using autonomous AI agents. The platform includes specialized agents such as HR, CTO, CFO, CEO, and Risk Management agents that collaborate intelligently to perform tasks like AI-driven interviews, candidate evaluation, operational analysis, workflow automation, and executive decision-making. Unlike traditional AI assistants or single-agent chatbots, Synapse AI focuses on collaborative intelligence where multiple AI agents communicate, reason, and coordinate together to solve complex organizational workflows in real time. The system supports multimodal interactions including text, documents, reports, and speech inputs, allowing users to simulate real enterprise environments and automate time-consuming operational processes. For example, users can conduct AI-powered HR interviews, upload business reports for executive analysis, or generate strategic recommendations through coordinated AI agent discussions. Technically, the platform is built using Next.js, FastAPI, Gemini AI, Speechmatics, Supabase, Docker, and Vultr cloud infrastructure. The architecture uses scalable distributed services, asynchronous processing, and modular AI orchestration to ensure reliability, low latency, and production-style deployment readiness. Synapse AI demonstrates how autonomous AI systems can function like real organizational teams, helping businesses improve operational efficiency, reduce repetitive manual work, accelerate decision-making, and create scalable intelligent enterprise workflows for the future of AI-driven organizations.

DemoMaster

DemoMaster

DemoMaster turns a repository or live demo URL into a polished narrated pitch video. It runs an agent workflow that understands the product, captures real browser evidence, writes a demo-led script, generates voice narration, and exports a final video with transcript and agent logs. In this demo, DemoMaster starts from a Sauce Demo URL, shows the agent run in progress, produces the pitch workspace, and plays the generated result video with real product footage. DemoMaster uses a Next.js web app with Playwright for real browser capture and video evidence. Gemini powers product understanding, pitch writing, capture-aligned narration, and voice generation. Featherless is used as an open-model critic for pitch quality checks, while Speechmatics verifies the generated narration against the transcript. The final video is rendered in-browser with canvas/WebM, then converted to MP4 with FFmpeg for submission.

ATRIO Boardroom

ATRIO Boardroom

**Founders and family offices decide alone.** Big calls get either delegated to a single advisor (fast, single point of failure) or convened with a committee (slow, hard to schedule, hard to audit). ATRIO Boardroom is the middle option: an AI boardroom that holds a real debate, enforces a per-tenant mandate at machine speed, and replays every decision in six months. ## Try the live demo **URL:** http://45.77.52.54:8080 (Vultr, Frankfurt) Click **Demo founder** on the sign-in screen — one click, no email — then type a boardroom question. Watch 5 specialist AI agents stream real Gemini 2.5 reasoning live, ~25 s. Go to Treasury, propose a SHV-xStock buy, try to self-second-sign (API refuses), open a new tab as **Demo CEO**, second-authorise, trade executes against Kraken paper. Download the board-pack PDF. Open the audit log. Six minutes, full lifecycle. ## The wedge - **Debate**, not consensus-on-rails. Six personas with distinct system prompts, distinct model assignments (Gemini 2.5 Flash for specialists, 2.5 Pro for Counsel), and dissent-driven turn-taking. - **Enforce**, at the API. A per-tenant `Mandate v1` (permitted instruments, daily loss limit, single-instrument max, permitted side) is the only path to a treasury action. Two-party auth cannot be bypassed by the UI. - **Audit**, by default. Every turn, vote, model call, and state transition writes to an append-only log. Exportable as JSONL + manifest. ## Why this isn't slideware - **381 / 381** backend tests pass at **90.68 %** coverage - **24 / 24** demo-video structural + **14 / 14** OCR verification - **54 / 54** pitch-deck structural + **12 / 12** OCR verification - **5 / 5** live multi-agent debate against real Gemini in ~25 s (no mocks) - **19** real bugs found and root-caused during the sprint ## Sponsors used Vultr · Google Gemini · Featherless · Speechmatics · Kraken xStocks · LiveKit. License: Apache 2.0.