Manufacturing & automation · Regenerative medicine
Cytera CellWorks, now Mytos
A London company that builds machines to grow human cells without a human touching them. It began as a lab automation hackathon, spent a year being written up as a cultivated meat play, and now runs its own automated factory inside the UK national cell therapy centre.
The flask that stays the same
Almost every automation company that has walked into a cell therapy cleanroom arrived carrying a robotic arm. It is impressive to watch and it has mostly not worked.
Mytos went the other way. Its machine, iDEM, is a box about the size of an under-desk fridge. The operator loads reagents and the same T flask a scientist would use by hand. Fluidics, pumps and valves do the coating, feeding, passaging, harvesting and imaging. A microscope sits inside, the images go to the cloud, and the scientist checks the culture from a phone.
The choice that matters most is the least glamorous one: the flask does not change, so a process worked out by hand can move onto the platform without redesigning the biology around a new vessel.
“We use exactly the same culture flasks that customers are growing cells in manually today,” Afshar told Labiotech in 2018. Eight years later the pitch has not moved.
What manual culture costs
Stem cell derived therapies are starting to work: dopamine neurons for Parkinson’s, islet cells for type 1 diabetes, retinal cells for vision loss. The bottleneck has moved from whether the biology works to whether anyone can make enough of it.
That making is done by hand, by people who are scarce and slow to train, inside cleanrooms that cost a fortune per square metre. Afshar puts the cycle at about a year before an operator is reliable, and says the good ones then leave. Variability is the other half: how hard you tap a flask and how you read confluence are judgement calls, and they show up in yield.
Two students and a hackathon
Ali Afshar was doing a PhD at Imperial College and building robots to get out of repetitive bench work. Ignacio Willats was running entrepreneurial events. They met at one, organised a lab automation hackathon, and found more demand in the room than they expected.
The company was registered on January 19, 2016 as HackScience Limited. The first plan was to automate the whole lab. They surveyed scientists and narrowed it to one job.
“We surveyed over 100 scientists from different companies and labs and they told us that [cell culture] was the number one thing that was taking their time to do manually,” Afshar said.
In the summer of 2018 they went through Y Combinator. That November the register read Cytera CellWorks Limited.
Three names, three markets
The record shows a company that has changed its market twice, and a story about itself edited to match.
When TechCrunch covered the YC batch in August 2018, it put Cytera in the cultivated meat business and described automation for growing turkey meat from a petri dish. Four months later Afshar told Labiotech the target was pharma R&D, and was already talking about cell therapy manufacturing.
By 2020 that was concrete: an Innovate UK co-funded project with Orchard Therapeutics. In February 2022 the register changed again, to Mytos Bio Limited. The 2023 Series A was still sold on R&D productivity, with heart cells that start beating after twelve days. Only in 2024 and 2025 did the company move squarely onto therapeutic manufacturing. Its own account is shorter: the about page says it started six years ago automating cells for US pharma R&D, and the cultivated meat chapter is not mentioned.
From selling boxes to running the factory
In January 2025 Aspen Neuroscience agreed to put iDEM into its 22,000 square foot Torrey Pines facility to make dopaminergic neuronal precursor cells for ANPD001, its autologous Parkinson’s therapy. Three partners followed that September: StemSight in Finland for corneal cells, Rinri Therapeutics in the UK for hearing loss, Novadip in Belgium for bone. Pluristyx, a Seattle iPSC line supplier, signed in December.
Then the model changed. In October 2025 Mytos stopped only selling equipment and became a manufacturer, opening its first contract development and manufacturing site inside the Cell and Gene Therapy Catapult’s Stevenage Manufacturing Innovation Centre. It targets clinical readiness by September 2026 and a stated ceiling of 1,500 autologous or 25,000 allogeneic doses a year. That is a different company from the one that sold benchtop boxes, and a much harder one. The deadline it set itself falls in the month this profile was compiled.
What is proven, and what is still claimed
| Evidence | What the record shows | Source type |
|---|---|---|
| Corporate history | Companies House 09957823: incorporated Jan 19, 2016 as HackScience Limited, renamed Cytera CellWorks Limited Nov 14, 2018, renamed Mytos Bio Limited Feb 3, 2022. Wholly owned by Mytos, Inc. of Delaware. | Public record |
| Founding year | Companies House says January 2016. TechCrunch in 2018 wrote “Founded in 2017 by two college students in the U.K.” The 2023 Series A release says “Founded in 2018.” | Claims differ |
| Patents | US12221600B2, automated cell culturing, priority Aug 19, 2018, granted Feb 11, 2025, inventor Ali Afshar, assignee Mytos Bio Limited. One further application, US20230087656A1. Two records in total. | Public record |
| Public funding | Innovate UK project 105852, March 2020 to March 2021, with Orchard Therapeutics (Europe) Limited. Cytera led it: project cost £766,136, grant offer £383,068. | Public record |
| UK finances, 2025 | Mytos Bio Limited lost £3,946,194 after £3,733,100 in 2024. Cash fell from £2,306,413 to £1,074,775. Net assets £393,354, accumulated losses £15,337,878, average employees 27. No profit and loss account delivered. | Public record |
| Revenue signal | No turnover figure is public. Trade debtors rose from £53,881 to £584,319 during 2025. | Public record |
| Named customers | Aspen Neuroscience, StemSight, Rinri Therapeutics, Novadip, Pluristyx. Five named partners in twelve months, no contract value disclosed. | Company and partner stated |
| The Aspen Phase 3 | Mytos said in April 2026 that Aspen is gearing up for Phase III this year, using iDEM. ClinicalTrials.gov lists one Aspen study, NCT06344026, a Phase 1/2a of nine patients, still enrolling by invitation, updated May 14, 2026. | Not on the registry |
| Capacity, papers, FDA | Stevenage is claimed at up to 1,500 autologous or 25,000 allogeneic doses a year. No clinical batch released yet, no PubMed hits, and no 510(k) or PMA under either name. | Not found |
Read plainly: the technology is real, patented and in paying hands, and the customer list improved sharply in 2025. What is not on any public record is a clinical batch released from Stevenage, a revenue number, or a registered late stage trial using cells that iDEM grew.
What to watch
- Whether Stevenage reaches clinical readiness in September 2026, the date Mytos set for itself a year earlier.
- A registered Phase 3 for Aspen’s ANPD001. The registry still shows only the Phase 1/2a.
- The 2026 accounts for Mytos Bio Limited, due by September 30, 2027. Cash at that entity more than halved during 2025.
- Any Form D after October 2023. The last one on file is nearly three years old.
- Whether a partner discloses a contract value, turning five logos into a revenue line.
In their words
“At the time, lab automation felt suboptimal”
Ali Afshar, co-founder, to TechCrunch, 2018 · Independent
“We surveyed over 100 scientists from different companies and labs and they told us that [cell culture] was the number one thing that was taking their time to do manually”
Ali Afshar, Labiotech interview, December 2018 · Interview
“We’ve even heard people say they can look at a cell culture and tell who in the company has been growing those cells, which is obviously not a good thing for consistency and producing reliable data”
Ali Afshar, Labiotech interview, December 2018 · Interview
“The most tedious process in biotech is growing human cells by hand. Every scientist needs to spend hours a day growing cells and coming in on weekends to look after them”
Ali Afshar, CEO, Series A release, November 2023 · Company release
“One company told us they were seeing failure rates as high as 50% in one of their differentiation processes due to manual variability”
Ali Afshar, Regen Report Q&A, April 2026 · Interview
“We’ve even seen up to 90% reductions in hands-on cleanroom time for some workflows”
Ignacio Willats, co-founder and CCO, Regen Report Q&A, April 2026 · Interview
“Regenerative medicine has extraordinary potential, but the cost and scalability of manual manufacture are barriers to reaching patients at scale”
Ali Afshar, CDMO launch release, October 2025 · Company release
Related companies
Sources
- Public recordMYTOS BIO LIMITED overview, previous names and filing history, 09957823
- Public recordFull accounts to 31 December 2025, Mytos Bio Limited
- Public recordFull accounts to 31 December 2024, Mytos Bio Limited
- Public recordForm D, Mytos, Inc., CIK 0001995930
- Public recordForm D, Mytos a series of TBD Angels Master LLC
- Public recordInnovate UK project 105852
- Public recordUS12221600B2 and US20230087656A1
- Public recordAspen Neuroscience registered studies, NCT06344026
- IndependentCytera CellWorks aims to bring cell culture automation to your dinner plate
- InterviewHow Automation Will Enable the Age of Cell Therapy
- Independent$19M Series A Funding for Mytos
- IndependentAspen Partners with Mytos to Automate Stem Cell Production
- IndependentMytos inks multiple cell therapy manufacturing partnerships
- IndependentMytos Launches Automated CDMO
- InterviewInside iDEM, Mytos’ automated platform
- CompanyMytos Announces $19M Series A Funding
- CompanyMytos Launches Automated CDMO for Stem-Cell Derived Therapies
- PartnerAspen Neuroscience advances automated production of ANPD001
- CompanyMytos site and Cytera CellWorks archives
Profile researched and written by Healthcare Discovery. Last updated September 29, 2026.
