Srihari Sritharan is an investment fellow at Kaleida Capital, one of the premier neurotech investment firms, founded by Ariane Tom. Kaleida is known for being neurotechnology operators and researchers before investors, which gives the firm an unusual edge on the business of BCI. It was for that insight that we spoke to Srihari.
Sri did his PhD at Penn under Timothy Lucas, building wireless somatosensory prosthetics for spinal cord injury patients. A motor implant can help you move an artificial hand, but it cannot tell you what you are holding. Srihari worked on the feedback loop, and produced the first chronic neural interface with the sensory nuclei in the brainstem.
To read the pressure on a fingertip without burying electronics in it, he built passive microfluidic antennas: tiny reservoirs of liquid gallium alloy injected into polydimethylsiloxane that fill when you press down, changing the antenna's length and shifting its frequency.
Then came deep research across nanophysics, stem cell work, and nanoparticles that respond to epilepsy EEG signals to deliver drugs, before he crossed over to the commercial side. At Acadia Pharmaceuticals, Sri ran diligence on a $967 million collaboration with Stoke Therapeutics on RNA therapies for rare neurodevelopmental disease, which is to say he learned how nine figures of neuro money actually moves. Today he runs business development at Beacon Biosignals, an EEG machine-learning company that sells quantitative brain-data endpoints into biopharma trials, and he does venture fellow work at Kaleida Capital on the side.
So I asked him: how do you get to a billion-dollar revenue business with BCI? He answers from a position where he understands both the research and how to commercialise it.
Why BCI has been difficult to fund
"The concern has always been that no one will agree to do such an invasive surgery, that is basically permanent, unless you are so impaired it's going to make a step change in your life," he said.
There is deep value in the hardware and the software, but never the explosive scale of a drug, or even of an established neuromodulation play like deep brain stimulation. So BCI sat at the very back of the pipeline, the moonshot the funds took, in his phrase, "just to do it."
AI changes the calculus for the tech and investors
Deep learning got good enough that you no longer have to be pressed against a neuron to read it. The old assumption was a dense bed of electrodes, hundreds or thousands, jammed as close to the tissue as possible, decoding the symphony of thought. Now the algorithms pull something meaningful out of far worse signals.
Srihari points to the aging chronic implants from the BrainGate era, arrays half walled off by scar tissue with many electrodes long dead, that can still move a cursor because the software caught up to the silence.
And the flip side of that coin is the unlock. "Maybe you never needed something that invasive in the first place," he said. EEG, EMG, fNIRS, whatever noisy signal you can get, becomes usable once you have enough of it. Cheaper to build, lower surgical risk, and suddenly the addressable market is something a fund can believe in.
Data is considerably more valuable now, because the algorithms have caught up.
He points to Synchron as the purest version of the bet. A stent slipped into a blood vessel, no contact with any neuron at all, a couple of dozen electrodes reading the neighbourhood and trusting that adjacent signal is enough. The money has followed.
But some problems still demand a dense array pressed right against the cortex, which is why the high-resolution players like Paradromics and Precision Neuroscience still matter, and why he has run diligence across both ends of that spectrum for Kaleida. "Not all deficits are the same," he said. Out at the far edge are the companies convinced that if you just record enough EEG you can read cognition itself, which he thinks is a stretch dressed in the language of large models.
Would healthy people line up for elective brain surgery?
He does not see it. "It's so high risk, and I don't even know if it's high reward yet."
Even a low-risk implant like an IUD sees modest uptake against a pill, because it is extremely painful on the way in and on the way out. The one device he could think of that people willingly keep inside them at all times is a continuous glucose monitor, and only because the alternative is worse.
Vagal nerve stimulation for chronic anxiety or PTSD might catch on, he allowed, because the population is enormous and the need is real. But the vision the excitable end reaches for, where a company like Forest Neurotech folds into a bet like Merge and everyone eventually volunteers to jack in, is the stretch.
Where adoption comes fast
It is the opposite of glamorous. People with genuine deficits: ALS, locked-in syndrome, the loss of speech. People who have nothing else.
"ALS drugs don't work," he said. Once an implant is provably safe and lasts thirty years instead of burning out in three, he expects that group to take it up quickly, because for them the cost of waiting is total.
When should a BCI founder call Kaleida?
Early, and before the story is fully baked. "We're aiming mostly at pre-seed, seed, and Series A," he said, the lower-valuation end where a team is still finding its feet. Kaleida's edge is that the bench is "almost entirely people who either did a PhD in BCI or are adjacent to it," some out of Synchron, some out of Meta's VR work.
Sri is open to hearing from founders at the early stages, and from investors who want the same exposure to the frontier from people who have done the science and built the businesses themselves.
References
- Srihari Sritharan on LinkedIn
- Kaleida Capital: kaleidacapital.com
- Beacon Biosignals: beacon.bio