Photo Courtesy of Fareed Salmon.
Madhav Lavakare’s (YC ’25) dive into the world of assistive technology traces back to his junior year of high school, when he witnessed a close friend grapple with hearing loss. After his friend, unable to understand what was being said in class, had to drop out of school, Lavakare embarked on a quest to explore existing solutions, only to encounter extravagant costs and imperfect outcomes. Traditional aids, such as hearing aids and cochlear implants, often cost tens of thousands of dollars and merely amplify sounds without clarifying them, so users may still easily miss out on what’s being said. Closed captioning on a device offers a solution to this issue of auditory comprehension but at the cost of missing visual communication cues. Hard-of-hearing individuals often have to play a “tennis match” of looking at subtitles on a mobile device and then back up at the person speaking. Since a simple phrase such as “Hey” may vary in meaning based on the expressed emotion and facial cues, closed captioning remains an imperfect solution for those who rely on it.
Madhav’s invention, TranscribeGlass, emerges as a solution to these two issues faced by the hard-of-hearing community.
Hoping to bridge the gap between auditory comprehension and visual cues, Lavakare designed TranscribeGlass to be affordable real-time captioning glasses. After interviewing and testing prototypes with members of the deaf community every weekend while working on this project full-time in India, Lavakare envisioned a device that seamlessly integrated real-time captions into the user’s field of view.
Using Bluetooth technology, TranscribeGlass transmits caption data to a hardware device attached to a pair of glasses, mirroring the functionality of a movie projector. Imagine a transparent screen that acts as a see-through projector, enabling users to effortlessly follow along with captions in real-time, whether in a cinema hall, a classroom, or in day-to-day conversations.
TranscribeGlass has evolved through a user-centric approach, driven by continuous feedback and iterative refinement. Lavakare refers to American entrepreneur Eric Ries’s mantra of the “build-measure-learn” cycle in describing how TranscribeGlass’s first prototype, albeit bulky and rudimentary, served as a starting point that was later refined by feedback. From India to Gallaudet University for the deaf and hard-of-hearing in Washington, D.C., over 350 individuals have been part of the prototype testing process, paving the way for subsequent iterations with greater compactness, longer battery life, and user-friendliness.
The culmination of Lavakare’s efforts is a sleek device weighing less than eighteen grams—allowing users to not feel the burden of the device on their glasses—and boasting an impressive eight-hour battery life. Unlike conventional alternatives entailing exorbitant costs and invasive surgical procedures, TranscribeGlass offers a plug-and-play solution at $95, democratizing access to assistive technology. Its light design and prolonged battery life ensure uninterrupted usage, transcending the constraints imposed by traditional aids.
Furthermore, Lavakare’s device offers a unique feature—caption-source independence. Unlike its counterparts, which often rely on a single provider and demand constant internet connectivity, TranscribeGlass utilizes a diverse array of methods to generate accurate captions, even in offline mode. By integrating with speech recognition software from companies such as Google, Apple, and Microsoft, as well as connecting to movie subtitle files and human transcription providers, it ensures accuracy and reliability across various contexts.
The impact of this technology resonates with users worldwide. After a video of TranscribeGlass’s product showcase went viral with over twenty-six million views, demand for the product’s final release has skyrocketed with over fourteen thousand sign-ups for preorders. “I had a woman write to me saying, ‘Hey, I have two deaf twin daughters who are starting their freshman year of college, and this device would change their lives,’ and another shared how her father’s recent hearing loss has isolated him, believing TranscribeGlass would bring a newfound sense of inclusion and confidence for him. So, the feedback we’ve gotten has been very positive,” Lavakare said.
Looking ahead, the future is ripe with possibilities for TranscribeGlass. Lavakare hopes to incorporate real-time translation into his model and cater to disabilities such as autism and ADHD by including visual cues and reminding users of past conversations. Lavakare’s journey with TranscribeGlass continues, fueled by his ability to innovate and driven by his commitment to ensure accessibility for all.