We’ve come up with some important facts that led to the creation of an AI-assisted Speaking Feedback tool:
1) Roughly 90% of English learners receive zero feedback on their speaking skills.
2) English teachers are not knowledgeable of industry-specific communication norms and can only listen to one person at a time.
3) Oral evaluations are difficult to control without third party influence.
Our app gives excellent feedback on any conversation with unlimited concurrent speakers, using a combination of industry-relevant knowledge sets, expert tips, enhanced voice recognition and the power of generative AI.
Using Synergetics Education’s unique approach, we empower language teachers, learners, and test invigilators to understand and immediately improve every aspect of English-speaking skills.
Version 1 can do the following:
- Record and Analyze: Easily record yourself speaking and let our advanced algorithms analyze your speech patterns, pronunciation, and grammar usage.
- Dual Transcripts: Receive two comprehensive transcripts for each recording—a faithful representation of your original speech, and a refined version that mimics native speaker fluency. Compare and learn from both to bridge the gap between your current proficiency and fluency.
- Expert Tips and Insights: Our team of experienced language specialists has curated a wealth of tips and insights to help you master advanced English usage. Discover the nuances of idiomatic expressions, refine your grammar skills, and gain a deeper understanding of the language.
- Progress Tracking: Stay motivated and track your progress over time. Witness your growth firsthand as you build confidence and fluency with every recording.
- Customized Learning Paths: Synergetics English Feedback tailors the learning experience to your unique needs. Whether you want to focus on specific areas of improvement or explore a wide range of language skills, our app adapts to your goals.
We wanted to make this a useful tool for teachers, especially those with large classes and minimal teacher aides. We used the best practices of English Language Teaching (ESOL, ELT, EFL, TESOL, CEFR, Cambridge), feedback from initial users, and UX design principles to make the Feedback App simple and effective. This includes new pathways to learning connected to cultural frames and the power of generative AI to create, evaluate, and adjust information. Thus we are creating a Synergy of language, content, and design personalized to each learner. Just click the button.
We adjusted the Best Time Range for 1-5 minutes of dialogue or monologue. This is because we want to promote speaking practice and gather enough information for the AI to notice any frequent errors. The App will pick up both speakers but it doesn’t make a distinction so it’s better to have each app open for each person. The feedback consists of an AI-assisted transcript analysis, reformulation, and tips. It is sequenced into three “Feedback Events”:
Event 1 Compare & Contrast: Learners can compare their own transcript with the “improved” transcript offered by generative AI. The color coding indicates “wrong” and “right” but that’s only a visual stimulus for the learner. What’s actually occurring is the highlighted words in the pink text are changed or altered, trying to keep the original meaning but changing the language to improve syntax and vocabulary range. It can change the meaning, as any interpretation can, but this is also a great feature for comparison by the learner or other reviewers of the transcripts.
Event 2 Analysis & Alternatives: The AI offers the learner specific problems and an alternative to say it “correctly,” or at least with more flow, clarity, and with more advanced vocabulary. What this inclusion of more complex vocabulary and structure does is ‘teach’ the learner new words and structures through reading and comparing their speech with one specifically made informative for language learning exercises.
Event 3 Clarity & Conclusion: the AI offers general English advice, which helps more advanced learners but also gives tutors and invigilators a benchmark for the learner. This provides invaluable information to anyone analyzing the learner’s progress, but also gives the learner direct and specific support for improving their general communication in English based on the spontaneous speech provided not by a pre-constructed dialogue or monologue.
Later, “past sessions” can be reviewed and analyzed both by the tutor or external reviewer, and AI can establish patterns ranging from the strengths to weaknesses of the learner’s speech and how to improve. A person could record ten practice speeches and try to fix errors each time, creating an iterative circle of improvement simply impossible before generative AI but now literally at the push of a button.
One use case of this App for courses at Synergetics is for our large “Pre-Conference Networking in English Experience” where groups of people going to Websummit can meet & mingle, practicing their English and Learning some Key Networking Phrases and Techniques. Each participant has five conversations to practice, and ends up with up to five personalized reports. The algorithms behind Speaking App are adjusted and trialed with the skills and theme of Networking, but all participants can have their own conversations and get immediate and relevant feedback to enhance their networking experience at the imminent event a week later where they’re immersed in a specialized type language, namely “tech” and “networking”.
The Speaking Feedback App is fit for classrooms where all students have laptops or mobile devices; in online classes where each student has their own device; or in small group tutorials and invigilations with one device. It’s suitable for all ages but particularly aligned to teens and adults with intermediate and above proficiency. We invite tutors, schools, individuals and companies to try out the Beta version before our launch at Websummit:
Sign up for the Networking in English Experience at any of the dates here.