Fluent: Turning thought into text with a brain-computer interface
The promise of brain-computer interfaces (BCIs) has long been the stuff of science fiction, but a Melbourne-based startup is turning it into reality. Meet Fluent, a TRAM Air 2025 venture developing a minimally invasive BCI that translates thought into text. Their goal: to revolutionise how we interact with digital systems and give a voice to those who can’t speak.
As AI becomes more integrated into our lives, the way we communicate with it remains surprisingly limited. Most digital interaction still relies on typing, clicking or speech, methods that can be clunky, inaccessible, or socially awkward in certain settings.
“What we’re trying to do is remove that friction entirely,” says Fluent CEO and Co-Founder Tim Mahoney. “Imagine being able to think something and have it instantly transmitted to a device. That’s the future we’re building.”
While the long-term vision is a seamless thought-to-AI connection for anyone, Fluent’s first focus is on people with communication impairments, those who are paralysed or living with neurological conditions that make speaking difficult or impossible.
A safer, accessible approach
At its core, Fluent’s device reads brainwaves, no surgery inside the skull required. Unlike some BCIs that involve high-risk implants, Fluent uses a minimally invasive approach, making the technology safer, faster and more accessible.
“Our device measures brainwaves and uses AI to convert them into text,” explains Co-Founder Dr Dean Freestone. “It’s not just about what’s technically possible; it’s about making this technology safe, usable and scalable in the real world. It sits outside the skull but underneath the skin, which dramatically lowers risk and cost.”
The device sits outside the skull but underneath the skin
Tim adds, “This is useful for everyone, but particularly for people with disabilities. It allows communication at a conversational pace for those who can’t speak, without the friction of typing or speaking aloud. Eventually, we believe it could even become an elective procedure for anyone wanting a faster, more seamless connection to AI.”
The technology draws on years of research, combining signal processing, neuroscience, and machine learning to decode user intent from brain activity. In simple terms, you think, and the device types.
From research to startup
Fluent emerged from two research groups that met through TRAM, one focused on decoding brain signals using AI, the other on the hardware to record them.
“Tim had been working on the signal acquisition side through his PhD, and we’d been exploring speech decoding from EEG signals,” says Dean. “When we realised Tim’s method offered a safe, high-quality way to measure the signals we needed, it was a lightbulb moment. We had the makings of something powerful.”
Today, the team includes experts from the University of Melbourne’s biomedical engineering department: Professor David Grayden, Associate Professor Sam John, Dr Farhad Goodarzy and Stephen Ruffels, alongside neurosurgeon Dr Andrew Morokoff, who is advising on device usability and surgical readiness.
“Between us, we’ve led some of Australia’s biggest medical technology startups,” Dean adds. “It’s a small core team, but the talent is exceptional and that’s what we need to go far.”
Fluent team - Dr Dean Freestone, CTO and Co-Founder, Farhad Goodarzy, CPO and Co-Founder, Tim Mahoney, CEO and Co-Founder, Associate Professor Sam John, Founding Inventor
Key milestones
Since forming, Fluent has achieved several critical early wins:
Filed a provisional patent on its core technology
Secured proof-of-concept funding from the University of Melbourne
Completed feasibility studies in animals, with human clinical trials next
Published foundational work on sub-scalp EEG, proving its safety and long-term viability
Brought on industry-leading mentors through TRAM, including Synchron CEO Tom Oxley, whose company is at the forefront of implantable BCIs.
Alongside these tangible milestones, the team has also experienced a shift in perspective that reinforced their belief in the venture’s potential. “There was a moment I realised this wasn’t a scientific problem anymore, it was an engineering challenge,” Dean recalls. “That’s when I knew we were onto something real.”
The path ahead
In the next 12 to 24 months, Fluent plans to run its first human studies, validate its decoding technology, and develop the systems needed for large-scale data collection and model training. They’re pursuing both private investment and grant funding to support this next phase.
“We’re aiming to show, with real human data, that we can decode speech from brain activity in a safe, scalable way,” says Tim. “From there, we build the product.”
Initial applications focus on people with neurological disorders or injuries, but longer-term, Fluent envisions broader accessibility. “We believe this could eventually become an elective procedure,” Dean says, “Not because people have to, but because it enhances how they engage with AI and digital systems. It’s about human capability.”
Clarifying the conversation
BCIs are often misunderstood or sensationalised. Fluent wants to change that narrative.
“We get asked: are you reading people’s thoughts? The answer is no,” says Tim. “We’re detecting intent to communicate nothing more, nothing less. And we’re designing everything with security, consent and safety at the core.”
TRAM’s impact and the future
Both founders credit TRAM for more than just practical knowledge: it helped build the team, align their vision and accelerate their transition from research to startup.
“It’s a steep learning curve,” adds Tim. “But having that support and being surrounded by people going through the same challenges made a huge difference. It gave me the confidence to go from PhD student to CEO.”
Tim Mahoney pitching at a 2024 TRAM event
“TRAM takes someone from knowing nothing about commercialisation to being part of an incredible team working on unbelievable technology,” Tim says, “From customer discovery to developing and validating a business model, it’s been transformative.”
Dean adds, “TRAM provides frameworks for alignment and execution. It doesn’t matter if you’re new to startups or experienced, it helps you improve and accelerate. We definitely wouldn’t be where we are without the program.”
The team plans to stay involved with TRAM, potentially mentoring future teams and continuing to collaborate from the ecosystem’s shared office space.
Fluent’s device could transform life for people who can’t speak due to illness or injury and unlock new ways of thinking about how we interact with machines.
Read more about the 2025 TRAM Air cohort
Get in touch with Masha to learn more about TRAM Air: masha.pelipas@unimelb.edu.au
Sign up to TRAM’s newsletter