Closing data gaps across R&D and the clinic

What we'd propose

• Digital CDMO

  Digital twin for AAV production optimisation

  We build a comprehensive virtual bioprocess model of the FLT201 and SPR301 production cycles, physics-based and data-driven, that runs virtual experiments to predict the impact of media formulation changes and scale-up parameters before committing to a wet-lab campaign or a CDMO tech transfer.

  • Virtual bioprocess simulation

    Predicting outcomes before the campaign runs

    Build models of transfection, cell culture and purification from historical campaign data so that the impact of parameter changes is known before the next run, rather than after it.

  • Full and empty capsid yield prediction

    Identifying the separation conditions before running chromatography

    Use the model to predict chromatography conditions that maximise full-capsid recovery, so the purification step starts from an informed starting point rather than from a default method.

  • Scale-up risk assessment

    Simulating CDMO performance before tech transfer is finalised

    Run the digital twin against the CDMO's expected equipment parameters to identify critical mismatches before manufacturing campaigns begin.

  • Wet-lab experimentation cost reduced by predicting parameter impact virtually before committing to a campaign.
  • Full-capsid yield improvement means more therapeutic doses per bioreactor run.
  • Tech transfer risk is reduced by simulating CDMO performance before equipment is committed.

  We've done this before

  • Creating a Digital Twin to Simulate Biopharma Processes
  • Automated Cell Line Control with Advanced PAT

  See how we do Digital CDMO
• Digital CDMO

  IT and OT connectivity for Stevenage and CDMO sites

  We connect the Stevenage pilot-scale OT environment — bioreactor sensors, chromatography systems and analytics — to a unified enterprise data layer using OPC UA, and establish data pipelines and shared data contracts with CDMO partner sites so both sides of the hybrid manufacturing model see the same batch data in the same format.

  • OPC UA connectivity for Stevenage OT

    Connecting bioreactors and chromatography to the enterprise layer

    Instrument bioreactor sensors — metabolites, pH, dissolved oxygen, cell density — via OPC UA so process values are readable by enterprise systems without proprietary middleware.

  • Cross-site data pipelines

    Shared batch data with CDMO partners

    Establish agreed data contracts and secure pipelines between Spur's enterprise systems and each CDMO partner site, so batch progress and quality data flows both ways without manual consolidation.

  • Unified manufacturing dashboard

    One view of every site

    Build a centralised monitoring view that brings Stevenage pilot-scale production and each CDMO site's commercial output into one dashboard, so the technical operations team has real-time visibility without logging into multiple systems.

  • Real-time visibility into batch progress at every site, without manual status reporting.
  • Reduced batch failure rates through proactive intervention when live data shows deviation.
  • Shorter tech transfer timelines because the CDMO receives a data contract, not just a document.

  We've done this before

  • Scaling Biotech Manufacturing via Universal IT/OT Architecture

  See how we do Digital CDMO
• Enterprise AI

  Clinical trial data lake for R&D and clinical operations

  We build a cloud-based data lake that consolidates R&D data from Stevenage — genomic, protein and assay data — with clinical outcome data from Boston, unified by an agreed ontology so that the two estates can be queried together and longitudinal analyses like lyso-Gb1 reduction over time can be produced without manual data reconciliation.

  • Ontology-driven data harmonisation

    One agreed set of terms for assay, specimen and result

    Define assay, specimen, result, instrument and lot as explicit entities with agreed relationships, so that a query written once runs across both the Stevenage and Boston data estates without translation.

  • Data pipelines from Stevenage instruments

    Automated flow from the lab to the lake

    Build ingestion pipelines from the Stevenage analytical instruments and LIMS, with schema validation so that bad data fails at the boundary rather than silently contaminating the lake.

  • Longitudinal patient outcome analytics

    Tracking treatment response over time

    Enable cohort analysis of lyso-Gb1 reductions, GCase enzyme activity and clinical endpoints, so that lab observations and clinical outcomes can be joined in one view for regulatory submissions and dose optimisation.

  • Lab-to-clinic feedback loop shortened from days or weeks to the time it takes to run the analysis.
  • Regulatory data packages assembled from the lake rather than compiled manually for each submission.
  • Bioinformatics bottlenecks dissolved: genomic and clinical data live in the same system.

  We've done this before

  • Ontology-Driven Ecosystem for Biomanufacturing Data
  • Creating a Digital Lab for a Global Biotech Player

  See how we do Enterprise AI
• Digital Lab

  Automated batch release QC workflow for Phase III

  We implement digital QC workflow automation that captures manufacturing data from bioreactors, purification systems and analytics directly into compliant electronic batch records, then applies rule-based QC review that approves in-specification batches automatically and flags only exceptions for specialist review.

  • Electronic batch record integration

    Manufacturing data captured at source

    Automate capture of process data from bioreactors, purification systems and analytics into structured electronic batch records, eliminating manual transcription and its associated transcription errors.

  • Automated QC review workflows

    Release by exception

    Implement rule-based QC review that automatically approves batches meeting specifications, flagging only exceptions for human specialist review — reducing the batch review queue without changing the compliance standard.

  • Audit-ready compliance package

    Regulatory documentation assembled automatically

    Generate automated compliance reports and audit trails meeting FDA and EMA requirements, reducing the preparation time for regulatory inspections and enabling inspection readiness on a rolling basis.

  • Batch release time reduced from weeks to days, directly supporting the Phase III production schedule.
  • Inspection readiness is a property of the system, not a冲刺 at the end of each quarter.
  • QC headcount scales with the number of batches through automation, not through hiring.

  We've done this before

  • Digital Transformation of QC Labs: From Paper-Based Legacy to Integrated LES
  • Transforming Raw Bioprocess Data into Actionable Insights

  See how we do Digital Lab
• Agents

  AI agents for gene therapy regulatory document work

  Narrow AI agents that handle the repetitive part of regulatory document assembly for FLT201 and SPR301 submissions: drafting CMC sections from source batch records, pulling together deviation and change-control summaries from LIMS and MES data, checking completeness against the Common Technical Document template before review begins, and finding every controlled document that a standards change touches. A named reviewer approves every output.

  • Drafting CMC regulatory sections from source records

    First drafts from batch data

    Generate the first draft of CMC regulatory sections — process description, characterisation data, batch data tables — directly from the source batch records, so technical writers edit and approve rather than assemble.

  • Deviation and change-control summary drafting

    From instrument data to reviewable document

    Pull deviation reports and change-control summaries from LIMS and MES data, checking completeness against the site's own checklist before the document enters the human review queue.

  • Change impact search across the document set

    Scope of a standards change established on day one

    When FDA guidance, EMA guideline or an internal specification changes, retrieve every controlled document that references it and rank them by how directly they are affected, so the update scope is known before the work begins.

  • CMC submission packages assembled faster because agents handle the assembly step.
  • Review queues move faster because documents arrive complete, not missing sections.
  • The scope of a guidance or standards change is established by search rather than by recollection.

  We've done this before

  • Process Transfer Reporting System

  See how we do Agents