☰

EVerZom

From laboratory turbulence to industrial exosome production

Industry
Biotechnology (Exosome Therapeutics)
Headquarters
Paris, France
Public information as of
March 2026

A4BEE prepared this analysis from publicly available sources. It reflects our own reading of EVerZom's published strategy and is not endorsed by, or produced in cooperation with, EVerZom. Company website

Strategic priorities

EVerZom has built its position around a turbulence-based method for generating extracellular vesicles, in which controlled fluid dynamics drive a massive exosome release from parent cells. The technique was developed at the Matière et Systèmes Complexes laboratory (Université Paris Cité and CNRS) and is now the company's central industrial asset. Total disclosed funding is over EUR 13 million, with a EUR 3 million France 2030 grant awarded in June 2025 specifically for industrialisation, and a EUR 10 million Series A closed in October 2025 with Capital Grand Est, the EIC Fund, Audacia and Aloe 6.

The clinical lead is EVerGel, a topical exosome therapy intended for inflammatory diseases, targeted to enter Phase 1/2 trials in 2026. Producing clinical-grade material at scale means moving the turbulence method from the 2-litre laboratory reactors used in research to 50-litre GMP reactors, with larger scales in view. Validation work is run through the EFS (Établissement Français du Sang) partnership, and the company has stated that the first successful transition from laboratory to clinical scale has already been demonstrated.

The second strategic pillar is licensing the turbulence platform to pharmaceutical partners and CDMOs (Contract Development and Manufacturing Organizations — firms that develop and produce biological materials on behalf of other companies). Co-development agreements with companies including GENFIT, and participation in the Drug Cell alliance, point to a model where EVerZom supplies both the manufacturing method and the data package that lets a partner reproduce it. Standardised digital interfaces become part of the product that is licensed.

Industrialisation on this timetable depends on three capabilities the company has identified explicitly: real-time monitoring of Critical Process Parameters and Critical Quality Attributes during the 50-litre runs; the ability to demonstrate ALCOA+ (the data integrity principles of Attributable, Legible, Contemporaneous, Original, Accurate, plus Complete, Consistent, Enduring and Available) and GAMP 5 (Good Automated Manufacturing Practice, a risk-based framework for validating automated systems in regulated production) compliant records ahead of regulatory inspection; and a digital recipe package that can travel to a partner site and be commissioned reliably.

Challenges we see

  • Operations Manufacturing

    Holding Kolmogorov-scale parameters steady at 50 litres

    EVerZom's turbulence-based exosome generation is governed by Kolmogorov-scale fluid dynamics, where system geometry, kinematic viscosity and impeller rotation speed together set the conditions for a clean exosome release. At the 50-litre scale, where the company is moving for clinical-grade production, those parameters cannot be held by manual adjustment alone.

    Where a turbulent-flow method is run by hand, the operating window is whatever a single operator can hold in mind. As the scale grows, the same window has to be held by the instrumentation and the control system rather than by the person at the reactor.

  • Digital Integration

    Connecting bioreactor sensors to analytical results

    Data from the four innovation blocks — mother-cell modification, exosome generation, engineering and formulation — lives in disconnected spreadsheets and paper records. Within exosome generation itself, the Operational Technology of the bioreactor (impeller speed, CO2, temperature) and the Information Technology of the analytical instruments (Nanoparticle Tracking Analysis, viable cell density) sit on separate systems.

    When process signals and analytical results live in different systems, the live state of a batch is reconstructed by hand at the end of the run. Bringing the two sides into the same time-ordered view means a deviation surfaces during the run rather than after the harvest.

  • Compliance Regulatory

    Defining CMC for a heterogeneous advanced biotherapy

    Exosomes are classed as advanced biotherapies, with FDA and EMA frameworks still being shaped. EVerZom must define CMC for an inherently heterogeneous extracellular vesicle product while meeting GMP and GAMP 5 requirements.

    A product whose active component is a population rather than a single molecule asks for evidence that the manufacturing process is what defines the population. Each batch release becomes a case for showing the process was reproducible, which only holds if the underlying data holds together end to end.

  • Operations Manufacturing

    Linking potency assays to manufacturing data

    Regulators require evidence that each EVerGel batch is biologically functional. Standardising potency assays requires an automated overlay of biological results with the manufacturing parameters that produced them, often described as Golden Batch comparison.

    Where potency data and manufacturing data live in separate files, the link between them is assembled at review time. Putting the link together as the data is generated turns a research question into a query, and makes a Golden Batch reference a living artefact rather than a static document.

  • Digital Regulatory

    Protecting proprietary methods as data flows widen

    EVerZom's value sits in its patented turbulence algorithms and proprietary cell modification protocols. As the company moves toward clinical trials and partnership models, industrial data starts to flow between R&D, manufacturing and CDMO partner sites.

    Where the boundary between internal and partner-facing systems is set ad hoc, the data path that carries the company's know-how is whatever the next integration happens to be. Designing that boundary as an explicit architecture, with segmentation, authentication and access control built in, makes confidentiality a property of the system rather than a matter of policy.

Opportunities, by urgency and business impact

Each bubble is one opportunity, numbered to match the list below. Further right means it bites sooner; higher means a bigger effect on the business. A bigger bubble means a bigger implementation effort.

Source: A4BEE analysis of public sources

  1. Reading bioreactor signals and exosome metrics in one view

    Today's EVerZom run is judged successful only after the release phase is complete, because Critical Process Parameters (impeller speed, shear stress, CO2, temperature) and the biological metrics that describe the batch (Nanoparticle Tracking Analysis, viable cell density, exosome size and concentration) are not seen together while the run is in progress.

    An Industrial Data Platform that brings OT sensor data and analytical IT results into a single time-ordered model gives scientists a live view of the run, a Golden Batch overlay, and a deviation alert that fires while the batch is still on the line. The company's CTO has called out the need to calculate the right micro-scale vortices and monitor quantity and size of exosomes released all along the vesiculation protocol — exactly what a unified live view enables.

    • EVerZom, Exosome generation
    • EVerZom press release, October 2025 (EUR 10M Series A)

  2. Bridging bioreactor control and laboratory IT with OPC UA

    The bioreactor control layer (impeller speeds, CO2, temperature) and the laboratory instruments that characterise the exosomes (Rochester, Hamilton analysers, Vi-CELL) speak different protocols, and the data reconciliation between them is done by hand for every batch.

    An OPC UA-based (Open Platform Communications Unified Architecture — a vendor-neutral industrial communication standard) connectivity layer on the bioreactor PLCs (Programmable Logic Controllers — the industrial computers that run reactor hardware) and analytical instruments carries both sides into the central data platform without disrupting production, so the same run is visible to the engineer, the analyst and the quality organisation at the same time.

    • EVerZom, Innovation platform
    • The Drug Cell alliance statement on automation, AI and digital manufacturing

  3. Putting GAMP 5 and 21 CFR Part 11 onto the exosome workflow

    The transition from experimental to industrial data integrity is a prerequisite for Phase 2/3 trials. Paper-based logging and fragmented spreadsheets for Critical Process Parameters and Critical Quality Attributes leave ALCOA+ gaps that surface during regulatory inspection.

    A Laboratory Execution System with GAMP 5 and FDA 21 CFR Part 11 (the US regulation governing electronic records and electronic signatures in regulated environments) compliance captures data at the instrument, applies electronic signatures, and produces a real-time view of audit readiness, so the inspection in 2026 finds evidence already in the record rather than a reconstruction.

    • EVerZom statement on cost-effective large-scale standardised production under GMP
    • Series A press release, October 2025

  4. Building a digital twin of the turbulent-flow bioreactor

    The Kolmogorov-equation-governed method requires precise mathematical control of fluid dynamics at increasing scales, and physical experimentation at 50 litres is expensive. Yield outcomes cannot be predicted before committing to a full production run.

    A digital twin that simulates Kolmogorov vortices, shear stress distribution and impeller dynamics lets the team optimise parameters virtually before committing to a physical batch, and provides a sandbox for AI model training that does not consume production capacity.

    • EVerZom, Exosome generation
    • EUR 2.5M EIC Accelerator award, 2021

  5. Packaging the turbulence method for partner sites

    Licensing the turbulence platform to CDMOs and pharmaceutical partners is a stated commercial pillar, but there is no standardised digital package that lets a partner reproduce the production process at their own site without months of bespoke engineering.

    A digital recipe and a Plug-and-Produce equipment interface — using MTP (Module Type Package — a standard for modular, vendor-neutral process equipment modules) where it fits — let EVerZom hand partners a transferable manufacturing package that configures bioreactor parameters at the receiving site and shortens technology transfer from months to weeks.

    • EVerZom, Partnering
    • GENFIT research collaboration announcement

What we'd propose

  • Enterprise AI

    Real-time process intelligence for the 50-litre reactors

    We connect the bioreactor's OT layer and the analytical instruments on the IT side to a single time-series platform, and we put a Golden Batch overlay and deviation detection in front of the scientists, so a 50-litre run is read as it happens rather than reconstructed afterwards.

    • Unified sensor and analytical dashboard

      One view of a running batch

      Bring temperature, CO2, impeller speed and shear stress together with Nanoparticle Tracking Analysis and viable cell density on one time-ordered dashboard, so scientists see both the physical state of the reactor and the biological state of the batch at the same moment.

    • Golden Batch overlay with live deviation alerts

      Comparing each run to a reference

      Plot the current run against a validated reference batch and alert when Kolmogorov-scale parameters drift outside the tolerance envelope, giving operators minutes to act rather than a finding in a later report.

    • Automated bio-KPI pipeline

      Metrics calculated from the data

      Compute growth rate, exosome concentration and size distribution automatically from raw sensor and analytical feeds, so the analyst opens a populated dashboard instead of building a spreadsheet for each batch.

    • Batch outcomes are read during the run, not after the harvest.
    • Engineers, analysts and quality work from the same live view of the same numbers.
    • Golden Batch comparison becomes a continuous practice rather than a quarterly exercise.

    We've done this before

    See how we do Enterprise AI
  • Digital CDMO

    OPC UA-based IT/OT convergence for exosome manufacturing

    A vendor-neutral connectivity framework that bridges bioreactor control systems and laboratory information management with OPC UA and network segmentation, so the data path from sensor to scientist dashboard is designed rather than improvised.

    • OPC UA integration layer

      Standardised machine-to-platform communication

      Deploy OPC UA servers on bioreactor PLCs and on analytical instruments so the same protocol carries OT and IT data into the central platform, and add new equipment against the same standard rather than a new driver.

    • Network segmentation between OT and IT

      A documented boundary between layers

      Design and implement VLAN (Virtual Local Area Network) segmentation that isolates the bioreactor network while permitting controlled, encrypted data flow to the platform, so cybersecurity is built in rather than added later.

    • Multi-vendor device integration

      Connecting the instruments already on site

      Build the connectors for Rochester, Hamilton, Vi-CELL and similar instruments that are already in use, so the current equipment fleet enters the unified platform without replacement.

    • Manual data transfer between bioreactor and lab is replaced by an automated data path.
    • Future equipment is added against an existing standard rather than a new integration project.
    • Proprietary algorithms sit behind a segmentation boundary that is documented and auditable.

    We've done this before

    See how we do Digital CDMO
  • Digital Lab

    GMP-compliant digital quality management for EVerGel

    An integrated Laboratory Execution System and electronic batch record platform with GAMP 5 and 21 CFR Part 11 compliance, so the EVerGel production workflow carries ALCOA+ evidence by construction and is inspection-ready for the 2026 clinical trial submission.

    • Automated instrument data capture

      Captured at the instrument, not typed in

      Connect analytical instruments to the Laboratory Execution System with barcode scanning and direct ingestion, so ALCOA+ data integrity is enforced at the point of collection instead of being reconstructed during review.

    • Electronic batch records with 21 CFR Part 11 signatures

      Paperless, signed production documentation

      Replace paper logbooks with electronic batch records that timestamp, attribute and validate every Critical Process Parameter entry, with electronic signatures that meet 21 CFR Part 11 from day one.

    • Real-time compliance dashboard

      Audit readiness as a live metric

      Surface data-integrity metrics, unsigned records and missing entries on a compliance dashboard, so the team knows the inspection posture continuously rather than during a pre-audit scramble.

    • Manual entry errors drop out of the production record at the source.
    • Audit readiness becomes a continuous property of the system rather than a project.
    • Batch release waits on the analytical result, not on the paperwork that follows it.

    We've done this before

    See how we do Digital Lab
  • Digital Lab

    Adaptive digital twin of the turbulent-flow bioreactor

    A simulation-driven digital twin of the Kolmogorov-scale fluid dynamics in EVerZom's turbulent-flow bioreactor, used to optimise impeller parameters virtually, predict yield outcomes before physical runs and generate synthetic data for AI model training.

    • Fluid dynamics simulation

      Modelling the Kolmogorov-scale vortices

      Build a high-fidelity model of the turbulent-flow bioreactor that simulates shear stress distribution, impeller dynamics and microcarrier behaviour at 50-litre scale, so parameter optimisation begins in software before any wet work.

    • Predictive yield modelling

      Forecasting batch outcomes

      Train machine learning models on historical batch data to predict exosome yield, size distribution and purity from input parameters, with confidence intervals that say how reliable the prediction is.

    • Synthetic data for AI training

      More data without more runs

      Generate synthetic datasets from the digital twin to expand the training set for process models, reducing the number of expensive physical runs needed to improve prediction accuracy.

    • Parameter candidates are screened in simulation before they reach the reactor.
    • Scale-up from 50 litres to the next commercial scale is a process transfer task rather than a re-learning task.
    • R&D cost per optimised parameter falls as synthetic data supplements wet-lab runs.

    We've done this before

    See how we do Digital Lab
  • Agents

    AI agents for CMC documentation and regulatory submissions

    Narrow, reviewable AI agents that take the repetitive part of regulatory documentation work for an exosome therapy entering the clinic: drafting CMC sections from source records, checking submissions against the regulator's template before they enter review, and tracing every controlled document a standards change touches. A named scientist approves every output.

    • CMC drafting from source records

      First drafts pulled from system data

      Generate the first draft of CMC sections, clinical study summaries and process descriptions directly from the underlying manufacturing and analytical records, so the author edits and judges rather than assembles the document from scratch.

    • Template and completeness check before review

      Gaps caught before the queue

      Run the document against the FDA and EMA submission templates and the site's own checklist, returning missing or inconsistent sections before the document enters the formal review queue.

    • Change impact search across the document set

      Which documents a change affects

      When a standard, method or specification changes, find every controlled document that references it and rank them by how directly they are affected, so the update scope is established by search rather than by recollection.

    • Regulatory submissions arrive at review complete, so review queues move faster.
    • The scope of a standards or method change is established by search rather than by memory.
    • Every draft is traceable to the source records it was generated from and signed off by a named scientist.

    We've done this before

    See how we do Agents

Digital maturity: today and target

Scored out of 100 across six dimensions. The target is what EVerZom's own published ambition implies — not a perfect score.

Source: A4BEE analysis of public sources
Process Automation 20 → 75
Bioreactor parameters at 50 litres are not yet held by closed-loop control; the transition from manual impeller adjustment to automated control is a stated requirement of the EUR 3M France 2030 industrialisation grant.
Data Integration 18 → 80
OT sensor data and IT analytical results sit in separate systems across the four innovation blocks; a unified data platform that connects them is described as a precondition for clinical-grade production.
Regulatory Digital Compliance 22 → 82
Paper-based logging and spreadsheet-based records do not yet meet ALCOA+ in production; GAMP 5 and 21 CFR Part 11 compliant electronic records are a stated prerequisite for the 2026 Phase 1/2 trial.
Predictive Analytics & AI 12 → 70
No live predictive model exists for yield or batch outcomes at 50-litre scale; a digital twin of the Kolmogorov-scale fluid dynamics is the planned route to prediction before physical runs.
Cybersecurity & IP Protection 20 → 70
Proprietary turbulence algorithms and cell modification protocols currently sit in systems without an explicit segmentation boundary; widening data flows to CDMO partners raise the value of an architecture-level answer.
Modular Manufacturing Readiness 10 → 65
No MTP-compliant interfaces exist yet on the manufacturing equipment, which limits how cleanly the process can be packaged for technology transfer to partner sites.

Check this yourself

Our Service Portal has free self-assessments and market comparisons. These are the ones that line up with what we've read above — no sales call required.

Think we've read this right?

Talk to us

Related reading

This is an independent analysis prepared by A4BEE from publicly available information as of March 2026. It reflects A4BEE's own interpretation and opinion, is not affiliated with, endorsed by, or verified with EVerZom, and may be incomplete or inaccurate. All company names and trademarks are the property of their respective owners. To request a correction or removal, contact [email protected].