Speak’s AI Technology
& Learning Solution
The fundamental reason why your English speaking skills haven’t improved for the past 10 years is simple. You just haven’t had enough opportunities to speak English with a native speaker. Speak is solving this problem with its own speech recognition AI technology.
The fundamental reason why your English speaking skills haven’t improved for the past 10 years is simple. You just haven’t had enough opportunities to speak English with a native speaker. Speak is solving this problem with its own speech recognition AI technology.
Speak’s Core Technology
Speak's AI technology provides the same experience as having a real conversation with a native speaker, and becomes more sophisticated every day by gathering the user’s voice data for more than 10,000 minutes every day.
Speak's AI technology provides the same experience as having a real conversation with a native speaker, and becomes more sophisticated every day by gathering the user’s voice data for more than 10,000 minutes every day.
%203-min.jpg)
Fast and accurate speech recognition model developed internally
Latency
Accuracy
Real-time
Speak’s Automatic Speech Recognition model is the optimal result created by combining the English speech data of native speakers and the English speech data of over 1 million Koreans.
Speak's speech recognition model, which accurately recognizes the English pronunciation of Koreans, implements the core functions of the product with a fast recognition speed of less than 0.1 seconds and a high accuracy of over 95%.
Latency
Accuracy
Real-time
Phoneme unit analysis
and real-time feedback
Speak's Phoneme Recognition technology has been trained to instantly convert your voice into phonetic symbols for phoneme-by-phone speech analysis.
The user's phonetic symbols and best phonetic symbols are recognized and collated in real time to provide accurate feedback to the user.
Optimized speaking experience with AI technology
Experience speaking more than 100 sentences in English in 20 minutes, completed with artificial intelligence technology and products.
The fast and accurate speech recognition performance and automatic recognition function that allows conversations to proceed without having to touch the screen enhances an immersive, natural conversation-like experience.
The Creators of Speak
Speak was created by founders from prestigious American universities such as Harvard, Stanford, and Yale. The Speak team of top experts in each field, including artificial intelligence, product development, and language education works in offices in San Francisco, Ljubljana, and Seoul.
Connor founded and successfully sold the education startup ‘Flashcards+’ at the age of 17. After dropping out of Harvard University, he went through Thiel Fellowship and Y Combinator, and became a founder of Speak.
Graduating from the University of Washington with three degrees at the age of 16, Andrew began his entrepreneurial career after dropping out of his PhD in Neuroscience at Stanford University and co-founding Speak, with Connor who met from Thiel Fellowship.
Gang is a Yale graduate and has a long career as a product analyst at major US startups such as Dropbox and Thumbtack. He is in charge of product analysis and development of Speak.
While at Harvard University, SJ was Connor's roommate. After graduating and working as a data analyst at a startup in New York, Connor offered SJ to join Speak. SJ is in charge of operations at Speak.
Connor founded and successfully sold the education startup ‘Flashcards+’ at the age of 17. After dropping out of Harvard University, he went through Thiel Fellowship and Y Combinator, and became a founder of Speak.
Graduating from the University of Washington with three degrees at the age of 16, Andrew began his entrepreneurial career after dropping out of his PhD in Neuroscience at Stanford University and co-founding Speak, with Connor who met from Thiel Fellowship.
Gang is a Yale graduate and has a long career as a product analyst at major US startups such as Dropbox and Thumbtack. He is in charge of product analysis and development of Speak.
While at Harvard University, SJ was Connor's roommate. After graduating and working as a data analyst at a startup in New York, Connor offered SJ to join Speak. SJ is in charge of operations at Speak.
English learning solution completed
with Silicon Valley’s technology
Learning a language is not just about reading texts or watching videos. Much like the way we learn our native language, it requires a lot of iterative input and output, and accurate correction. Speak develops all the necessary functions and provides unlimited access to make this happen.
Learning a language is not just about reading texts or watching videos. Much like the way we learn our native language, it requires a lot of iterative input and output, and accurate correction.
Speak develops all the necessary functions and provides unlimited access to make this happen.
The Three-Step
Learning Method
Each lesson consists of three stages: “Today's Lesson,” “Speaking Practice,” and “Conversation Simulator.”
Through each course, you will be able to say more than 100 sentences aloud in English in 20 minutes per class, allowing yourself to make expression you’ve learned your own without knowing it.
Repetitive Patterns To Learn Expressions Rather Than Sentences
Speak's “Smart Review” automatically saves learned English patterns to review items and provides review classes.
The “interval iteration” learning algorithm guides reviews to be performed at the optimal timing, helping users remember the expression.
Most English education services provide a review function. However, it is only repeating the learned sentences so it can be memorized. But it’s difficult to apply those sentences in real life. For those who have struggled with this problem, the smart review function will be of great help.
AI Pronunciation Coach
Speak's pronunciation coach feature analyzes your pronunciation by phoneme, tells you how you pronounced it and how you should pronounce it, and provides the correct model pronunciation.
Forming Learning Habits Is Key!
Forming a consistent study habit is the key to effectively improving English proficiency. Speak constantly strives to provide an effective function for learning itself, as well as motivation to keep learning without giving up.
Firestreaks accumulate each consecutive day of finishing a lesson. Once you reach 100 consecutive days, you get a chance to win a prize!
There's nothing more motivating than a reward! Take part in our monthly challenge and win a free iPad.
Register the time you want to receive notifications to help remind you to study English.
Try the Habit Package and get a 100% refund when you complete the mission. The AI coach will help you achieve your mission.
