- PreSeed Now
- Posts
- 13 new deep tech startups you need to know
13 new deep tech startups you need to know
Here's what happened at the Conception X demo day...
I’ve been looking forward to today’s edition for quite some time, because I love a good demo day roundup.
My first was a Seedcamp event way back in about 2010, and the tech world has moved on a LOT since then. So what did this week’s Conception X demo day have to offer? Let’s find out…
But first: something that could be very useful if you’re starting something new… 👇 [Ad]
Free Notion and Unlimited AI
Thousands of startups use Notion as a connected workspace to create and share docs, take notes, manage projects, and organize knowledge—all in one place. We’re offering 3 months of new Plus plans + unlimited AI (worth up to $3,000)! To redeem the Notion for Startups offer:
Submit an application using our custom link and select Beehiiv on the partner list.
Include our partner key, STARTUP4110P67801.
13 brand-new deep tech startups you need to know about
Over the past seven years, London-based Conception X has been involved in the creation of 140 deep tech startups, and Riam Kanso, founder and CEO of the organisation told a packed room at 22 Bishopsgate in London on Tuesday that those companies have a combined valuation of around £500 million.
The crowd had assembled to meet teams from Conception X’s latest cohort of PhD researchers turning their work into deep tech startups. Of 108 teams, 13 ended up as finalists, pitching on the day.
Fun fact: someone involved with Conception X’s frustration at the lack of media outlets covering early-stage deep tech startups was part of the reason I started PreSeed Now, so I was very happy to be at the demo day to see first-hand what is coming down the pipe.
Here’s who pitched, with some initial thoughts on each. We’ll do deep dives on some of these startups in the coming weeks and months…
Hychor
They say: “Hychor is developing electrolyser technology to produce hydrogen directly from seawater, removing dependence on freshwater and energy-intensive desalination.”
We say: Green hydrogen is an active space for startups at the moment, as demand for this energy source increases. Hychor’s pitch quoted there being a £100 billion market for electrolysers that split water into hydrogen and oxygen.
Sea water needs to be desalinated to work with most of these devices. That’s annoying, as hydrogen production requires a lot of water, and, well, the world has increasing amounts of sea water.
Hychor aims to be an R&D company, licensing out its tech.
We previously covered Latent Drive which wants to do a similar thing in a different way, and we’ll be looking at another startup in this space soon.
From: University of Aberdeen
FTNanoAD
They say: “FTNanoAD has developed advanced, biocompatible nanoparticles that enable the targeting of brain tissue for drug delivery – to treat neurological conditions such as Alzheimer's disease, migraine, brain cancers and more.”
We say: FTNanoAD has a technique for delivering drugs directly into the brain via nanoparticles to treat neurological disorders in a more effective way than has previously been possible. The startup’s Joana Loureiro says this patent-pending tech is safe for longterm use.
While some other companies around the world can do a similar thing, FTNanoAD say they’re the only team with the ability to treat Alzheimer’s via this approach. The startup is building up towards clinical trials, planned for 2027.
From: University of Porto, Portugal
Mater-AI
They say: “Mater-AI is reducing waste heat by leveraging physics-informed and AI-accelerated thermoelectric materials discovery.”
We say: This team’s pitch put it that 70% of global energy is lost to heat, and waste heat is worth $152 billion. This heat can be harnessed for energy production using thermoelectric materials, but discovering such materials can take a long time.
Mater-AI has completed a proof-of-concept harnessing AI to cut down discovery time “from decades to minutes”. This could help to address the power needs of data centres, which generate a lot of heat they could be putting to practical use.
You can read more about this one in Tuesday’s edition of PreSeed Now next week!
From: University of Warwick
eteRNAbiologics
They say: “Eterna Biologics uses a proprietary CRISPR system to target disease-causing RNAs, achieving twice the efficacy of current RNA therapeutics, and accelerates development by combining CRISPR with single-cell biology to eliminate inefficient target optimisation processes, while creating a predictive engine for discovering novel RNA targets.”
We say: As the capitalisation of their name suggests, eteRNAbiologics (or just Eterna Biologics) is focused on RNA, with its CRISPR-lock platform technology for single dose genetic therapy.
Their pitch discussed using the tech to treat conditions such as blindness due to wet age-related macular degeneration, pushing genetic therapy forward. The startup is moving towards market readiness
From: University of Oxford
Hydra Manufacturing
They say: “Hydra has developed novel manufacturing hardware powered by deep learning for high-precision 3D printing of difficult-to-process materials, starting with advanced ceramics for aerospace.”
We say: Hydra Manufacturing’s pitch told us ceramic 3D printing would be a $4.2 billion market by 2032. The problem? Ceramic 3D printers are complex, inflexible, and expensive, so Hydra manufacturing has developed a better one.
They have achieved their first sale, to a ceramic manufacturing company, and are exploring metal 3D printing to diversify their product line.
From: University of Leeds
SynthLab
They say: “Building deep learning models to predict the behaviour of small molecule drugs in the body and accelerate the drug discovery pipeline.”
We say: This startup pitched remotely via Zoom from Boston in the US. SynthLab is looking to speed up drug discovery by generating data to predict chemical reactions without time-consuming experiments. To do this, they’re using neural networks.
They’re targeting drug companies that want to to predict how molecules will behave in the body. Having already raised a $190,000 pre-seed round, they’re currently looking to arrange design partnerships with biotech companies.
From: University of Oxford
Digial-Histo
They say: “Digital-Histo has developed AI software to transform breast cancer diagnostics, enabling earlier, faster and more accurate detection, and tailored patient care.”
We say: Digital-Histo wants to fix what they say are serious problems in breast cancer care: under-diagnosis and over-diagnosis. Both are costly, and this startup sees a gap in the market for AI-based tools to help reduce treatment costs.
After beginning with the NHS, they plan to target the US and EU, but first they need to optimise models and then validate the tech, which they’ll be doing with an NHS trust in London, as they move towards taking this to market.
From: Queen Mary University of London
Puncta
They say: “Puncta is building the first visual AI suite that unifies text-to-image generation and image editing in a single model, delivering 3x faster performance with professional-level quality and unprecedented cost efficiency.”
We say: Keep an eye on this one. With a web app and API, Puncta plans to take on giants like Adobe with text-to-image generation and image generation combined into one small, fast model.
Their proof of concept, launching soon, is an image upscaler that can do the job in seconds, as opposed to minutes with competing tech. Like similar generative AI companies, they plan to offer enterprise licenses and credits-based subscriptions, with an eye on both B2B and B2C markets.
From: Imperial College London
Puncta pitches at the Conception X demo day
Graphene Trace
They say: “Graphene Trace automates pressure ulcer prevention by continuous pressure mapping, powered by flexible graphene electronics and AI.”
We say: This one won the audience vote for best pitch. Pressure sores aren’t just a (literal) pain for the patients who get them, they take up valuable time and money that healthcare professionals could be putting to use elsewhere.
Originating in research from the same university that discovered graphene, Graphene Trace has developed a flexible fabric pressure monitor that can be embedded into wheelchairs and hospital beds to help avoid pressure sores.
The product is designed for continuous, widespread use, rather than one-off use in specialist clinics, like competing products. Having validated the tech through the iCURE programme, they’re now developing it from a prototype to an MVP that is expected in 2026.
From: University of Manchester
Omnibio
They say: “Omnibio creates organoid intelligence applications to create energy, data and space efficient solutions for the distributed sensor and robotics sector.”
We say: Why develop neuromorphic computing–AI inspired by the structure of the brain–when you can build a brain to run AI on instead?
Omnibio sees “organoid intelligence” as the next generation of AI. They’ve developed a 3D culture of cells that mimics the human brain.
As one co-founder said onstage (memorably in a West Midlands accent, which isn’t heard enough in the tech world) “AI is a little bit rubbish” when it comes to energy and data efficiency. And with foundational model developers running out of data to train conventional models on, this startup sees organic cells as a way to keep pushing progress in AI forward.
They see use cases for their tech in autonomous robotic control (e.g., for autonomous security monitoring), sensor data processing, and control systems.
This startup is also doing the Techstars and Creative Destruction Lab programmes.
From: University of Bath
SynthFairAI
They say: “SynthFairAI creates high-quality synthetic datasets reflecting real-world populations to test AI models for hidden bias more efficiently, ensuring enterprises comply with evolving regulations.”
We say: Varsha Ramineni began her pitch with the tale of how David Heinemeier Hansson was given 20x the credit limit of his wife for their Apple Cards, despite his wife having a better credit score.
SynthFairAI wants to avoid this kind of bias in AI. Ramineni, a DeepMind scholar, wants to leverage data from trusted sources such as the census quantify bias and help certify trustworthy AI.
The startup aims to develop a continuous testing solution for ongoing monitoring, and is looking to work with AI companies that want to increase their responsible AI credentials.
From: University College London
QMatter
They say: “QMatter accelerates drug design with quantum-inspired molecular simulation tools that are compatible with both conventional high-performance and quantum computing.”
We say: Like many in the quantum computing space, QMatter is looking to ‘quantum inspired’ tech for near-term market opportunities. Quantum computers just aren’t capable enough yet for a lot of the things people want to do with them.
QMatter’s tools run on classical computers. The startup wants to grow the computer-aided drug design market, and they says the techniques can be easily ported to quantum computers in the future.
This team came first in IBM’s Quantum Open Science Prize competition last year.
From: University College London
ReactWise
They say: “ReactWise has built a no-code AI co-pilot to accelerate and automate chemical process development for drug manufacturing, achieving up to 30x faster identification of optimal processes.”
We say: This S24 cohort Y Combinator startup is another team working in the drug development space.
ReactWise is developing high-quality datasets to train foundational models, and they say their experiments found a reduction the required experiments by up to 95%. They have secured seven paid pilots, including one with Pfizer.
From: University of Cambridge
Applications are now open for Conception X’s next cohort. You can find out more on its website.
Back on Tuesday
Want to know more about Mater-AI? We’ve got an in-depth profile of them for you coming in Tuesday’s edition of this newsletter, and we’ll check in on some of the others listed above soon.