The founding team has been involved in the development of Siri and Amazon’s smart voice assistant Alexa
According to foreign media reports , Finnish voice interaction technology developer” Speechly “recently completed a € 2 million seed round financing. This round of financing was led by Cherry Ventures in Berlin, followed by Seedcamp, Quantum Angels, Joyance Partners, Social Starts, and Tiny.vc. The financing will be used for technology research and development and the release of general-purpose API products.
Speechly was established in Helsinki in 2016. It mainly provides development tools for voice interaction functions for various application developers such as App. The founding team has been involved in developing Siri and Amazon’s smart voice assistant Alexa. Speechly’s voice interaction development tools can be used in e-commerce, games, digital medical, VR and other fields, and provide SDK toolkits for iOS, Android, Web and Unreal engines.
Speechly’s voice interaction technology is applied to shopping and food calorie counting apps
Google, Amazon and Microsoft provide developers with technical services for voice development. However, when developers use such third-party service platforms, they need to integrate their products into the platform ecosystem, and their brand exposure and operating data are also in the hands of the platform. “SpeechlyThe team believes that speech is a function, not an application, and should be integrated directly into existing applications. Based on the “Speechly” tool, developers can integrate voice interaction functions directly into the product, and they can maintain initiative on their own brand and data. At the same time, Speechly has also improved the timeliness of speech recognition and multi-modal fusion of speech and text.
-
Speech real-time interaction Speechly combines speech recognition with natural language understanding. It means listening and understanding at the same time. In this way, users can directly recognize the semantics when speaking, so as to realize real-time interaction. At present, when most tools on the market recognize speech, they need to convert the user ’s complete audio track into text and then understand the semantics. The timeliness cannot be synchronized in real time.
-
Multimodal fusion It is a combination of interactive methods such as speech and text. In some application scenarios, users do not have to use voice interaction throughout. For example, when ordering meals, Speechly allows the user to manually select the menu and only enter the address by voice.
In the field of e-commerce, Finland’s largest joint grocery store S-Group has used Speechly’s tools to develop a voice-driven iOS shopping app “S-Ostoslista”. Its voice-driven approach has reduced shopping time by 5-10 times. In the industrial field, based on Speechly’s tools, factories can create voice-controlled workflows, where workers can control machines from a distance using voice. In addition, areas such as ordering meals, customer service, medical orders, and financial services can be improved on speech-driven and semantic recognition functions based on Speechl’s tools.
Speech recognition technology, as a branch of artificial intelligence, is also a key area for various domestic technology companies. Bamboo Intelligence Developed a brain-like dialogue robot based on speech, semantics, and image interaction technology, which can act as a virtual customer assistant , Virtual enterprise assistant, etc. 云 知 声 applies technologies such as voice perception, cognition, and expression to the field of Internet of Things. Medical, financial, etc. launched smart solutionsAnd has reached cooperation with 20,000 enterprises including Geely, Midea and Ping An.