Our team recently attended the BioProcess International and Cell & Gene Therapy International conferences at Biotech Week Boston, which runs alongside, meeting with scientists, process development leads, manufacturing teams, founders, and investors.
The mood was enthusiastic and optimistic, anchored in the strong conviction that advances in cell and gene therapies will change how many diseases are treated. This is despite the manufacturing, data, and operational challenges that still stand between the promising science and patients.
Throughout the conference, a handful of themes came up consistently. Here are the five that stood out most:
1. Accuracy: automation is more than just productivity
Robotics and liquid handlers drew a lot of interest. Productivity is an obvious and important factor. But in this field, accuracy is just as important—and perhaps even more so—since even a single mistake in a lab process can have a significant downstream human cost.
Many of the sessions at Cell & Gene Therapy International stressed introducing automation early, because adding it after commercial approval carries the additional burden of proving results haven't changed. Software that orchestrates automated steps and captures data along the way was a recurring theme.
2. Consolidation: teams want more connected systems
The range of processes in cell and gene therapy is broad, and numerous discussions centered around the desire for platforms that can consolidate these processes efficiently and accurately. Today, many labs are still operating in the highly inefficient reality where they run an instrument, download the data, and then re-enter it by hand in another system. Despite the inefficiency and potential for error, multiple booth visitors described this as a normal part of their workflows.
Distributed networks of CROs and CDMOs complicate the picture, and tech transfer between partners remains an ongoing issue to be resolved. One CEO who came to us looking for a QC LIMS commented how much his organization would gain from running QC and other areas of the business on one connected system, without re-entering or reconciling data between teams.
3. Timing: when is the right time to go digital?
Several conference attendees noted that paper and spreadsheets often still work fine in early phases. However, as their programs near the end of the clinical stage and they prepare for regulatory review, teams inevitably start looking for the right moment to introduce digital systems. Factors that affect the decision usually revolve around implementation timeline and complexity, cost, resource availability, and the risk of a failed project.
Given this, it's understandable that many teams hold off on investing until a therapy has shown it can be commercially viable. In our view, the right time usually comes earlier, before the manufacturing process is locked for late-stage trials. Once a process is validated, every change has to be justified and documented, so systems introduced afterward carry the added burden of proving nothing changed. In fact, several sessions urged teams to bring in automation early for the same reason.
There are also practical signs that your team has reached that point:
- - instrument data is being re-entered by hand;
- - more batches, products or sites make spreadsheets hard to reconcile;
- - a tech transfer to a CDMO may be coming up soon;
- - pulling records together for an audit or investigation also means chasing information across systems.
Several companies we met were expanding and already finding that their data practices wouldn't scale with them. A platform that can adapt as the process evolves, without heavy IT or revalidation effort, makes an earlier start much more feasible.
4. AI: interest is high, and the path to value is becoming clearer
As at many recent conferences, AI was a major topic of conversation. Most of the sessions we heard centered on using AI to analyze data, though we noted that there was far less attention on how that data is captured. This gap matters because AI analysis is only as reliable as the data behind it – and reinforces the importance of the data layer and the need for capturing data with its context and relationships intact before applying AI to it.
Still, some vendors expressed frustration that the AI tools in their platforms aren't getting used. They added these capabilities because customers consistently asked whether their platforms “had AI” but then haven’t seen much adoption. This raises the question of how much real value these capabilities are delivering in the real world.
From our experience, AI tends to prove its value fastest in specific tasks, such as turning a plain-language process description into a configured workflow while people keep full control of the result. It also helps to choose systems that make data readable by any AI tool, so labs aren't limited to the features built into a single platform.
5. Realism: optimism balanced with patience
Overall, conference leaders are deeply optimistic about what these therapies can do for patients, and equally aware of how complex it is to bring one to market. That keeps investment expectations realistic and attention is focused on solving the most immediate bottlenecks.
In fact, this was reflected in Robert Hofmeister's keynote on moving CAR T therapy from a complex ex vivo process to an in vivo approach. While it generated plenty of excitement, the dominant discussion around it acknowledged it as a longer-term goal while teams keep refining ex vivo manufacturing.
Where Labbit fits
Cell and gene therapy covers a lot of ground, and Labbit focuses on managing a wide landscape of laboratory processes and the data around them across a therapy's lifecycle, from development through commercial manufacturing. Here’s how Labbit fits to each of the above observations:
Accuracy: Labbit integrates directly with lab instruments, so results flow into workflows without manual transcription. Each step is captured with its context as it happens.
Consolidation: Beyond QC testing, Labbit manages operational work such as orders, location tracking, and inventory. It also integrates with adjacent software to move data in both directions and keep it consistent across systems.
Timing: Labbit is designed to evolve from clinical to commercial scale, with workflow changes made through configuration instead of custom development. That keeps IT effort, vendor services, and revalidation to a minimum, which lowers the risk of adopting a LIMS earlier.
AI: Labbit captures data in a knowledge graph with full context, so it's readable by people and by the AI tools teams already use. Its AI configuration assistant turns plain-language process descriptions into BPMN workflows while users stay in control.
Realism: Labbit is built for complex science and iterative process development, so workflows can change as the process matures. As a knowledge graph-native, workflow-first platform, Labbit connects instruments, samples, materials, and results in a single system, with intuitive BPMN workflows that one booth visitor called "the way to do this." Data is captured with full context as work happens, so it's ready for people and for the AI tools teams already use.
If your team is weighing when to bring digital systems into your process, book a demo and we'll walk you through it.
Curious to learn more? Join our upcoming webinar, presented in partnership with Astrix, “Why Graph Architecture Is the Ideal Foundation for LIMS in GMP Manufacturing” where we’ll explore this topic in more detail.
