“Move fast but follow the rules scrupulously” — more or less the unspoken motto of every life sciences startup. Getting from idea to research to product-on-the-market has to be a fast timeline, but zero regulatory corners may be cut, and IP must stay secure.
Tissium arose from the labs of MIT, where a new polymer showed promise in revolutionizing the field of human tissue reconstruction. Leveraging that research, Tissium began developing biomorphic, programmable polymers to reconstruct damaged tissue and replace traditional stitches or other penetrating surgical techniques during surgery.
The startup relies on Box as a governed content layer for collaboration, workflows, and AI. Franck-Aurèle Medegan, Director, ESG & Finance at TISSIUM, oversees the financial and IT resources required to propel the company forward. He speaks to the high stakes of data, both structured and unstructured, in life sciences: “Trust, for us, is a prerequisite, because we are responsible at the end of the day for people's lives. Every platform we use has to be vetted.”
By centralizing its research and operational content on Box, Tissium’s teams are now able to take advantage of AI and automation without compromising governance.
Key takeaways:
- Tissium supports regulation workflows and compliance requirements with Box as the content layer
- Secure collaboration, both internal and external, is more logical with content consolidated onto a centralized platform with a clear, logical permissions architecture
- With archived R&D data segregated into a Box Hub, and specific permissions applied, the company can take advantage of institutional knowledge
Navigating strict regulatory pathways and 21 CFR Part 11 compliance
For medical device companies, innovation is only half the battle. “Regulatory pathways are critical,” Medegan explains. “You can build an incredible new medical product, but if you can’t get it through the regulatory process, you can’t commercialize it.”
As a French-based company, Tissium is subject to EU medical device regulation (MDR) requirements and audits involving European notified bodies. As a governed content layer, Box supports EU-regulated life sciences environments with built-in compliance for GDPR and GxP, enabling secure collaboration and compliant workflows aligned with EU MDR requirements.
Because it also operates in the commercial US market, Tissium must satisfy further requirements. “The FDA doesn't ask the same things as in the EU,” notes Medegan. “So we have to work hard to sort everything through and answer specifically to each geography.”
As a governed content layer, Box supports 21 CFR Part 11 readiness in the US, empowering Tissium to maintain compliance with FDA regulations. Medegan highlights Box’s comprehensive compliance documentation and dedicated compliance portal, which streamlined computer system validation using the GAMP 5 framework.
“Now,” Medegan says, “we can build automation securely knowing that if we are audited, it won't be a problem.”
Overcoming the roadblock of scattered data
Before Box, Tissium used a traditional hierarchical structure to manage content in legacy systems. Content was hard to find, and access and permissions hard to manage. “The main roadblock we faced was the way our data was structured, because we just let the contents and the folders sprawl,” explains Medegan. “It was frankly a mess. We knew that we could not go forward if we didn’t structure our data in the right way.”
The main roadblock we faced was the way our data was structured. It was a mess, and we knew that we could not go forward if we didn’t structure our data in the right way.
Partnering with Box Consulting, Tissium consolidated all of its content onto a centralized platform with a clear, logical permissions architecture. Box Consulting first met with teams across the organization to evaluate how they used content, what access they required, and what was shared externally. Having done this preliminary investigative work, Medegan recalls, “We felt comfortable from both a security and collaboration standpoint. Then, we deployed it overnight. It was a long night, but in the morning, everything was in the right place.”
Breaking down the organizational structure the company adopted:
- Automated classification:Box classification assesses and tags confidential files, helping protect Tissium’s intellectual property — the company’s most valuable asset
- Archiving: Older data is placed in dedicated archive folders to isolate active content queries and prevent people from surfacing irrelevant files, but also keep those files available for targeted search on institutional knowledge
- Controlled external collaboration: Externally shareable content is isolated in its own folder with specific permissions, lowering the risk of employees sharing unauthorized content, and Tissium uses Box Hubs as a curated destination for approved external sharing
“We’re leveraging more and more of the features of Box to keep permissions to a minimum, and it helps a lot,” Medegan says. “Your output is only as good as your input. Spending time cleaning the input is the most effective way to get reliable output.”
That brings us to AI.
10 years of institutional knowledge put to work with AI
For Tissium, a key motivation to experiment with AI was the massive volume of structured and unstructured R&D data that had accumulated over years of clinical studies.
“We think there are answers to questions we have today that we already answered ten years ago, and not exploiting that would be a major risk,” says Medegan. “We’re a startup, so we’re always at risk of losing time and money. Knowing what we have, we had to invest in AI.”
By leveraging the Box MCP server, Tissium will be able to provide AI agents with permissions-aware access to Box content. And Box APIs can support workflow integrations.
With these tools, Tissium is experimenting with integrating AI into workflows, automating document storage and querying. However, Medegan emphasizes that AI serves strictly as an assistant, alleviating cumbersome tasks like translation and study retrieval, rather than an autonomous creator. There’s pure human accountability where human lives are involved.
Painfully slow batch record processes replaced with automation
When it comes to compliant product documentation, Tissium’s process relies on handwritten batch records, which often span up to 20 separate files of 5 to 15 pages each for every single production batch. Using Box Automate, Tissium has launched a proof of concept to digitize the batch record creation process from initial production decision to clean room printing.
What took us the longest time was designing the process. Once we knew that, implementation was super fast and the responsiveness of the Box team was so helpful along the way.
Traditionally, life sciences rely on electronic batch record (ERB) software for this process, but switching to EBR software would have required approximately three years of implementation, extensive deployment and maintenance costs, and complex FDA validation. Box provided a much faster alternative while still maintaining legal compliance, and teams got the proof of concept up and running in just a few weeks.
“What took us the longest time was designing the process — what we actually want,” says Medegan, “But once we knew that, implementation was super fast and the responsiveness of the Box team was so helpful along the way.”
Automating batch records stands to be a huge time saver for Tissium. Where one batch typically requires 15 minutes of time, in proof of concept, automation brings that down to 2 minutes. The time savings adds up quickly when you consider the number of batches being run.
By uniting strict regulatory compliance, intelligent automation, and secure collaboration on a single platform, Box empowers Tissium to spend less time managing data and more time advancing the next generation of life-saving tissue reconstruction.
Read more about Box Governance and Box for life sciences.
