According to Zifo, biopharma companies must break a recurring pattern of disconnected experiments, fragmented evidence, manual reconciliation, and repeated context reconstruction that slows learning across the scientific value chain.
A new position paper calls for Design-Make-Test-Analyse workflows to evolve from linear, sequential processes into a connected network where scientists can do what they need, when they need to; scientific knowledge is captured, and context travels with every example, result, and decision.
Zifo describes the pattern as the ‘DMTA doom loop’, a flaw in the Design-Make-Test-Analyse process in which the scientific process is forced into a linear execution model, instead of adapting and changing based on what is required.
In its position paper, The Billion-Dollar Bottleneck: Why CMC is Strangling the Biopharma Pipeline, Zifo calls for the DMTA process to evolve from a linear sequence of handoffs into a connected scientific network where steps can be done in any order, repeated as required, and where the scientific context travels with each sample, result, and decision.
When Every Experiment Begins with Reconstruction
DTMA should form an interconnected learning-cycle network. Each experimental result should add to the organisation’s understanding of the relationship between materials, process conditions, equipment, analytical outcomes, product quality, and manufacturing performance.
However, in several CMC environments, evidence is distributed across scientific applications, instruments, spreadsheets, reports, PDFs, batch records, presentations, and individual expertise. Scientists may need to locate data, reconcile sample identifiers, reconstruct experimental histories, and recover the reasoning behind past decisions before interpreting results or designing the next experiment.
The position paper outlines that the underlying problem is not just fragmented data, but also the loss of scientific context and decision knowledge as data passes between systems, teams, and lifecycle stages. When results are separated from their samples, methods, process conditions, and decision history, the DMTA cycle can continue operationally but without functioning effectively.
Zifo believes a failure of scientific user experience sits at the core of the broken DMTA process. The biopharma industry has engineered IT ecosystems to satisfy retrospective compliance and data architecture needs, while making scientists’ daily workflows more difficult. To transform CMC and science, the enterprise must rethink scientific user experience.
Faster Handoffs Do Not Necessarily Create Faster Learning
The position paper advises against addressing the problem by digitising existing workflows or adding another isolated application.
Moving an inefficient process from paper or spreadsheets into a grid digital template does not necessarily deliver transformational impact, but it can produce a faster version of the same fragmented process while encouraging scientists to create workarounds outside governed systems.
Breaking the typical sequential process requires reimagining the current DMTA sequential process as an end-to-end scientific workflow in a connected network.
Under the proposed model:
- Scientists and engineers can move and jump between steps as required.
- Relevant historical data and evidence are available when designing experiments.
- Experimental intent and process context remain connected to samples.
- Test results retain their relationship to methods, conditions, and materials.
- Analytical outputs can be interpreted alongside prior evidence.
- Decisions are captured and remain traceable to the results and reasoning behind them.
- New knowledge becomes available to subsequent experiments and downstream teams.
These principles reflect the paper’s proposed shift from isolated systems toward connected scientific context and from a linear baton pass toward an agile DMTA network.
An Orchestration Layer Centred on the Scientist
Instead of proposing a single monolithic platform, the position paper calls for an orchestration layer that connects scientific activity across existing systems of record.
This environment should adapt to scientific process without replacing validated core systems. It should combine the evidence and context required for a task, allow information to be captured and reused at any step in the DMTA network, and reduce the burden on scientists of navigating applications and manually reconstructing relationships.
Reimagining the scientist user experience (UX and UI) means abandoning bloated, all-encompassing, rigid linear workflow templates in favour of small, modular, adaptable functional components based on critical capabilities adapted to scientists’ needs. The scientist requires a digital canvas that supports the immediate workflow in front of them while adapting as their science demands evolve.
Additionally, the position paper proposes governed Scientific Language Models (ScLM) grounded in proprietary scientific evidence. Used in a connected and appropriately governed environment, these models could help organisations capture and interact with their scientific knowledge while maintaining traceability to the underlying evidence.
From Completing Experiments to Compounding Knowledge
A connected DMTA network changes the objective from moving an experiment through separate functional stages to increasing organisational knowledge with each experiment or test.
This requires collaboration across CMC, process development, analytical development, quality, manufacturing, IT, data, and scientific informatics, as well as organisations prioritising the scientific and operational choke points where reconnecting evidence and context can produce meaningful value.
The position paper recommends targeted integrations and lighthouse implementations rather than a wholesale replacement of the existing technology estate, enabling organisations to begin with high-value workflows and expand the model within existing brownfield environments.
A scientist should not be forced through every stage if it is scientifically unnecessary. If contextual data already exists in the foundational layer, the researcher could move directly from Design to Analyse. Alternatively, they can move from Analysis into wet-lab execution because the underlying knowledge base is already contextualised and available. This fluidity breaks the linear delays of the traditional value chain and lets scientists navigate the lifecycle as the biological problem demands.
To download the paper, please click here: https://zifo.com/cmc-million-dollar-bottleneck-biopharma/
