Summary
- The problem
- A higher titre on a comparison plot is an observation, not a reason to choose a process. The conditions, material lineage, samples and uncertainty behind it sit in separate notebooks, spreadsheets and instrument files, and the reasoning is lost by the time the process reaches scale-up or transfer.
- Seal’s approach
- Experiments, actual execution and analytical evidence stay connected to a versioned process definition. Neil compares runs, identifies missing context and prepares the next experiment; scientists review the design and run the work.
- What changes
- Confounded comparisons stay visible instead of being read as an optimum, the next study is proposed for scientists to review before anything runs, and an unresolved scale question travels with the candidate to MSAT.
- Where to start
- One development decision in front of the team, with the runs, samples and results behind it. Book a demo.
Choose the next process from its evidence.
A development decision rests on more than the best number on a plot. Before a process goes forward, the team needs to know which factors changed between runs, what scale and materials each run used, and what the results cannot yet show. In this fictional example, Neil compares four runs, separates what each comparison supports from what it does not, and drafts the next study for the scientist to review.
Why teams choose Seal for process development
Process development is usually recorded in notebooks, instrument software and spreadsheets, then handed to manufacturing as a report that is re-entered as a batch record. In Seal runs, materials, samples and methods are linked records, so Neil compares runs from what they actually used, and the process the team chooses can become an executable protocol that keeps its development evidence.
PD-024 is a candidate. It produced 2.6 g/L, but feed and temperature both changed from the reference run, so this result does not isolate either factor. Each record below shows the conditions the run actually used.
View all values
| Record | Titre (g/L) | Viability (%) |
|---|---|---|
| PD-021 | 1.8 | 95 |
| PD-022 | 2.4 | 91 |
| PD-023 | 2.3 | 88 |
| PD-024 | 2.6 | 94 |
Lower temperature
PD-024
- Titre
- 2.6 g/L
- Viability
- 94%
- Working volume
- 2 L
- Feed
- 6 g/L/h
- Temperature
- 35 °C
PD-021 against PD-022: both use 2 L and 37 °C, while feed changes from 4 to 6 g/L/h. Titre rises from 1.8 to 2.4 g/L; viability falls from 95% to 91%. That is the closest comparison in this set. With one run per condition and incomplete lot and method context, I would not call it a demonstrated feed effect.
I’ve separated the two changed factors into a four-condition study. The proposal keeps the run records, measurements and open design decisions together, ready for the scientist to review.
PD-025 / next-study proposal
DraftSeparate the feed and temperature effects before selecting the process.
“We have been delighted with the rigour and agility shown by Seal in providing Ingenza with a robust GxP system that fully supports our process development and biomanufacturing operations.”
Keep each discipline’s evidence on the same process.
Upstream, downstream, analytical and characterisation teams each produce evidence about the same candidate. When their runs, materials and measurements are linked to one process version, a method change or a missing sample shows up in the comparison rather than after the decision.
Upstream development
Run comparison and proposed follow-up experiments.
Upstream development
Run comparison and proposed follow-up experiments.
Compare feed, culture conditions and actual run profiles without losing vessel, material or sample context.
Explore upstream developmentDownstream development
Step-recovery assessment with sample and pool lineage.
Downstream development
Step-recovery assessment with sample and pool lineage.
Follow recovery and quality across unit operations, pools and fractions. Locate the loss before choosing what to change.
Explore downstream developmentAnalytical development
A comparison that exposes method changes and missing context.
Analytical development
A comparison that exposes method changes and missing context.
Keep method versions, standards and sample preparation attached to the result used in a process decision.
Explore analytical developmentProcess characterisation
A study package with evidence and assumptions for review.
Process characterisation
A study package with evidence and assumptions for review.
Connect the study question, parameter ranges, responses and analysis to the proposed process understanding.
Explore process characterisationWork across the systems you already use.
Connect permitted records from your ELN, LIMS, historian, ERP and documents—or start with supplied data. Neil brings the relevant evidence into the work in Seal.
Procedures and quality records
Samples and analytical results
Materials and production context
Development and transfer documentsCarry the open scale question with the candidate.
Carry the evidence and the unresolved question together. In this study, the strongest result and the larger-scale run used different conditions.
PD-024
- Working volume
- 2 L
- Feed
- 6 g/L/h
- Temperature
- 35 °C
PD-023
- Working volume
- 20 L
- Feed
- 6 g/L/h
- Temperature
- 37 °C
Does the 35 °C condition hold up at 20 L?
These runs do not answer that yet. Keep the scale question with the candidate, so MSAT can assess the evidence and plan the next work.
See how MSAT carries the process forwardSee how the comparison is prepared.
3 minAQuestions and answers
How is this different from an ELN?
An ELN records experiments. This blueprint also connects the process definition to structured runs, actual conditions, samples, results and assessed claims. The transfer package keeps that context, giving the receiving team evidence and a starting definition rather than an approved manufacturing recipe.
Can development scientists still work flexibly?
Yes. The controls are configurable. In early development, scientists record data in a structured framework without being constrained by it. As the process matures, the controls can tighten to match its stage.
How does tech transfer work?
Transfer the selected definition with its run evidence, parameter rationale, equipment requirements and unresolved gaps. The receiving site maps its equipment and instructions, verifies the controls and approves the intended recipe. Preserve links to the development versions.
What about existing development data?
Legacy development data can be imported and linked to process definitions. New development work is structured from the start, and legacy data provides historical context.
How does this integrate with GMP manufacturing?
Development and manufacturing share linked definitions and evidence in Seal, so a manufacturing recipe can reference the runs behind each parameter. Development results do not by themselves authorise manufacturing execution.
Can this be used for GxP development work?
Yes, where your organisation defines the intended use and qualifies the configured workflow. Exploratory and controlled studies can use different permissions, review and verification. Compliance depends on that assessment and your procedures, not on the software alone.
Does promotion create a GMP recipe?
No. Promotion supplies a governed starting definition and its evidence, so nothing is retyped. Manufacturing still configures the site-specific recipe, equipment mapping, instructions and controls, and manufacturing and quality approve it before use.
What evidence gates characterisation readiness?
The team defines the criteria. Typical ones are a selected candidate, defined unit operations and parameters, representative materials and analytics, a scale-model rationale and a named owner for each open knowledge gap.
How are failed experiments used?
They stay in the process evidence with their conditions, results and conclusions. A later risk assessment, operating range or investigation can draw on them, and the team does not repeat a dead end.
How is this different from process characterisation?
Process development explores and selects the candidate. Characterisation then quantifies how parameters and materials affect performance and quality attributes, to support a control strategy. Both work from the same candidate, runs, parameters and samples in Seal.