☰

Afyren

Continuous bio-based acid production, now industrially validated

Industry
Industrial biotechnology (bio-based carboxylic acids)
Headquarters
Lyon, France
Public information as of
February 2026

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

Strategic priorities

Afyren runs the AFYNERIE® fermentation process across roughly 300 raw materials and 2 million hours of laboratory fermentation time, producing seven short-chain carboxylic acids at the AFYREN NEOXY plant in Carling Saint-Avold. The first plant entered continuous production in 2025 and recorded its first significant revenue in early 2026, with a target of 3 production units and 70,000 tons of annual capacity by 2028.

The early industrial ramp exposed post-fermentation work — separation and purification — that did not initially hold continuous operating conditions. Production was paused for two voluntary shutdowns during 2025 to address bottlenecks, and the company ended the year at 400 tons versus an initial single-digit million-euro revenue target. Capital activity in support of the ramp continued: a €23 million capital increase in November 2025 with Kemin Industries and Bpifrance, plus sustainability-linked loan drawdowns totalling €17 million through January 2026.

Beyond NEOXY, Afyren has signed a joint venture with Mitr Phol Group in Thailand to replicate the model on sugar cane feedstock rather than sugar beet co-products. R&D and IT/OT architecture decisions taken for that second site over the next few quarters will shape cost and stability for a decade.

As a sustainability-linked financing recipient, the company is aligning with the European Corporate Sustainability Reporting Directive (CSRD) and managing 20 steering and 6 strategic indicators across sites, alongside FSSC 22000, GMP+, Kosher and Halal certifications. Verified, continuous emissions data sits on the critical path for both the financing terms and the audits.

Challenges we see

  • Operations Manufacturing

    Stabilizing continuous operation at the separation and purification stages

    Some post-fermentation stages at AFYREN NEOXY have historically not reached the level of performance expected to enable continuous operation, and two voluntary shutdowns in 2025 limited capacity between July and October. Production ended 2025 at 400 tons and the plant recorded its first significant revenue in early 2026.

    Where purification is monitored as an end-of-batch result, the deviation signal arrives after the work has already been spent; reading process signals on the running batch narrows that window to the units that actually deviated.

  • Digital Integration

    Bridging the air-gapped IT and OT networks without compromising security

    Afyren maintains physical separation between its industrial OT network at Carling and its office IT network, citing CEO fraud, phishing and hacking as named risks that could paralyze production or compromise R&D data.

    A one-way controlled path between the two networks lets production data reach analytics without providing a return path into the control layer, which is the boundary condition that any cloud AI or remote supervision work has to respect.

  • Digital Integration

    Connecting Clermont-Ferrand fermentation data to Carling production parameters

    Two million hours of laboratory fermentation data have been logged across 300 raw materials. The 2025 bottlenecks indicate that data abundance has not yet translated into the parametric signals the plant needs.

    Where lab results and plant parameters live in different models, mapping them onto one shared schema is what makes a historical comparison across lab and plant possible without manual reconciliation.

  • Compliance Regulatory

    Collecting Scope 1, 2 and 3 emissions data continuously across sites

    Afyren's CSRD alignment covers 20 steering performance indicators and 6 strategic global indicators across sites, and the credit margin on its sustainability-linked loans adjusts with verified environmental impact. The plant also operates under FSSC 22000, GMP+, Kosher and Halal certifications.

    Manual quarterly collection puts the sustainability-linked loan terms and the certification audits on the same review cycle as the data gathering; sensor-level capture shortens that cycle and makes the indicators provable on demand.

  • Operations Manufacturing

    Designing the Thailand greenfield to avoid repeating the first plant's stability issues

    The Mitr Phol joint venture will replicate the AFYNERIE® model in Thailand on sugar cane feedstock rather than sugar beet co-products. Mitr Phol has appointed an EVP of Digital and Technology Transformation and is investing in Azure Data Factory and IoT integration.

    Standards for protocols, network segmentation and equipment data defined before procurement arrive with the machines instead of being assembled after handover, which is the difference between a designed and a retrofitted second plant.

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 separation and purification signals while the batch is running

    Post-fermentation stages at NEOXY have not held continuous operating conditions, and the bottleneck work has been driven by end-of-batch results rather than process signals during the run.

    Instrumenting the separation columns and downstream unit operations and streaming their process data into one monitored model lets a deviation be flagged against the running batch instead of after it, and gives operations and quality the same view of the run.

    • AFYREN, Business update on the AFYREN NEOXY plant, December 2023
    • AFYREN NEOXY: First Significant Revenue and Growing Production in 2025, January 2026

  2. Securing one-way flow from the OT network to the enterprise data lake

    Physical separation between the industrial OT network and the office IT network blocks cloud analytics, digital twin work and remote supervision, while cyber risks named by management (CEO fraud, phishing, hacking) rule out a less controlled path.

    A Zero Trust gateway with hardware-enforced one-way data flow, identity-based access control and a destination cloud data lake gives Lyon headquarters and the R&D team real-time production visibility without exposing the control layer to the office network.

    • Afyren, Annual Financial Report 2023
    • Afyren, 2024 Full-Year Financial Results press release

  3. Putting Clermont-Ferrand fermentation data on the same model as Carling production

    Two million hours of fermentation data covering 300 raw materials sit alongside Carling plant data, but the two describe experiments and industrial runs in different structures that are not joined together.

    An ontology-based data platform with explicit entities for feedstock, fermentation run, separation batch, lot and KPI, fed by automated pipelines, makes a Golden Batch comparison and a predictive feedstock model feasible rather than a manual exercise.

    • Afyren, 2024 Sustainability Report (2 million hour milestone)
    • Afyren, Industrial network architecture disclosure

  4. Capturing Scope 1, 2 and 3 emissions at the source for CSRD and the certification audits

    Afyren tracks 20 steering and 6 strategic indicators across sites for CSRD and the margin on its sustainability-linked loans, with the data currently assembled manually across FSSC 22000, GMP+, Kosher and Halal audit cycles.

    Sensor-level capture of energy, material and waste flows feeding a real-time dashboard brings the ALCOA+ (Attributable, Legible, Contemporaneous, Original, Accurate) data principles and the audit trail under the same source as the operations data.

    • Afyren, 2024 Full-Year Financial Results (sustainability-linked loan terms)
    • Afyren, 2024 Sustainability Report

  5. Specifying digital architecture for the Thailand joint venture before procurement

    The Mitr Phol joint venture commits to a second industrial site on a different feedstock, with decisions on equipment, network and IT/OT boundaries being taken over the next few quarters against a partner that has its own Azure data and IoT roadmap.

    Setting the protocol standards, network segmentation and equipment data requirements, with OPC UA (Open Platform Communications Unified Architecture) information models, before procurement means interoperability is a purchase condition rather than an integration project after handover.

    • Afyren, 2024 Full-Year Financial Results (Thailand JV disclosure)
    • Mitr Phol Group, digital transformation disclosures

What we'd propose

  • Digital CDMO

    Real-time process intelligence for NEOXY's separation and purification

    We instrument the post-fermentation stages at NEOXY with OPC UA connectivity and stream their process values into one monitored model, so a deviation is flagged against the batch that is currently running rather than assembled at the end of it.

    • Separation and purification data acquisition

      Bringing the column off the screen

      Connect controllers, sensors and analyzers on the separation and purification units through OPC UA (Open Platform Communications Unified Architecture) or MQTT (a lightweight publish/subscribe protocol widely used in industrial IoT (Internet of Things)) so process values leave the equipment in a documented, vendor-neutral form rather than staying inside a closed controller.

    • KPI calculation against a running batch

      Twelve-plus bioprocess metrics, live

      Automate calculation of the critical parameters — yield, purity proxy, residence time, energy intensity — within the data platform so the operations team sees the same numbers the engineers do, in minutes rather than at shift handover.

    • Golden Batch overlay on historical profiles

      Comparing the current run to the best one

      Overlay the current production run against the optimal historical profile for that unit operation so deviations from the expected trajectory are visible to operators while the batch is still being produced.

    • Deviations surface against the batch that is running, not against the one that already shipped.
    • One set of process data serves operations, quality and engineering instead of three separate extracts.
    • The release record for the unit operation is generated continuously rather than assembled at the end of the campaign.

    We've done this before

    See how we do Digital CDMO
  • Enterprise AI

    Zero Trust gateway between the NEOXY OT network and the enterprise data lake

    An architecture and reference implementation for moving production data out of the air-gapped OT network into an enterprise data lake under Zero Trust (an identity-first security model that verifies every request rather than trusting the network), keeping the control layer isolated from any return path.

    • Hardware-enforced one-way data flow

      Production data leaves, commands do not

      Deploy a data diode (a hardware device that allows traffic to flow in only one direction) or an equivalent unidirectional gateway so process data reaches the data lake while no command or session can return into the OT network, satisfying the cybersecurity baseline management already names.

    • Zero Trust access control for every data request

      Identity checked on every call

      Implement identity-based, just-in-time access for every read of production data, so researchers and executives reach the data lake on their own credentials with no standing privilege on the OT side.

    • Cloud data lake and dashboards on top

      From edge analytics to enterprise view

      Stand up a cloud data lake for the production stream, expose the Golden Batch and KPI views through dashboards, and give Lyon and Clermont-Ferrand real-time visibility into Carling without re-architecting the plant.

    • Cloud analytics and digital twin work can use real production data without lifting the air gap.
    • The cybersecurity baseline management already commits to is preserved by construction.
    • Leadership, R&D and operations work from the same numbers rather than from three separately maintained extracts.

    We've done this before

    See how we do Enterprise AI
  • Digital Lab

    Ontology-based data platform joining Clermont-Ferrand and Carling data

    A semantic data platform that defines once the entities both sites share — feedstock, fermentation run, separation batch, lot, KPI — and loads both laboratory output and production output against that model.

    • Shared bioprocess ontology

      One agreed set of terms

      Define feedstock, fermentation run, separation batch, lot and KPI as explicit entities with relationships, so a query written once returns comparable answers across Clermont-Ferrand and Carling instead of two dialects of the same table.

    • Automated pipelines from lab and plant

      From spreadsheet to source data

      Replace manual spreadsheet workflows with instrument-level ingestion, transformation and lineage capture for both the fermentation laboratory and the NEOXY plant, with schema validation at the boundary so bad records fail loudly.

    • Predictive feedstock and Golden Batch modelling

      From descriptive to prospective

      Train models on the joined dataset so the team can score a new biomass family against historical fermentation behaviour and overlay a current Carling run against its best historical match.

    • Two million hours of fermentation history become searchable against the plant's current runs.
    • New feedstock candidates for Thailand can be scored against the historical record before a pilot run is commissioned.
    • Integration work is done once against a shared model rather than once per point-to-point interface.

    We've done this before

    See how we do Digital Lab
  • Digital CDMO

    Digital architecture for the Thailand joint venture

    An architecture and standards package for the Mitr Phol second site: protocol choices, network segmentation and equipment data requirements agreed before procurement, so interoperability is bought with the equipment instead of built after handover.

    • Reference architecture for the greenfield plant

      One documented data path

      Specify how equipment, line supervision, the manufacturing execution system (MES, the software layer that tracks and orchestrates production) and the partner's Azure estate connect, including the segmentation model, so every supplier builds toward the same target.

    • Equipment data contracts in procurement

      What each machine must publish

      Write OPC UA information models and MQTT topic structures into procurement requirements, so interoperability and the data contract are purchase conditions rather than an integration project after handover.

    • Segmentation and cybersecurity baseline to IEC 62443

      Security zones from day one

      Define zones, conduits and remote-access rules against IEC 62443 (the international standard for cybersecurity in industrial automation) before commissioning, so the plant's security posture is the same on day one of operation as the audit assumes.

    • The patterns set here are reusable for any further greenfield unit Afyren adds after Thailand.
    • Interoperability is a property of the design instead of a property of the integration phase.
    • The risk of repeating the NEOXY ramp-up curve is reduced because the data architecture is no longer in flight.

    We've done this before

    See how we do Digital CDMO
  • Agents

    AI agents for CSRD and certification reporting work

    Narrow, reviewable agents that take the repetitive part of CSRD reporting, sustainability-linked loan verification and multi-standard certification tracking: drafting the indicator sections from source records, checking completeness against the framework, and finding every controlled document a change touches. A named person approves each output.

    • Drafting indicator sections from source data

      First drafts from system records

      Generate a first draft of a CSRD indicator or a sustainability-linked loan verification section directly from the underlying production and energy records, so the CSR team edits and judges rather than assembles.

    • Template and framework completeness check

      Gaps found before review

      Check a submitted report or audit packet against the relevant framework (CSRD, FSSC 22000, GMP+, Kosher, Halal) and the site's own checklist, returning missing or inconsistent sections before the document enters human review.

    • Change impact search across the document set

      Which documents a change touches

      When a standard, regulation or feedstock specification changes, retrieve 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 recollection.

    • Review queues for CSRD, certification and loan verification move faster because documents arrive complete.
    • The scope of a standard or feedstock change is established by search rather than by who happens to remember.
    • Every output is traceable to the source records it came from and signed off by a named reviewer.

    We've done this before

    See how we do Agents

Where QB Systems fits

Alongside our services we build QB Systems, hardware and software for bioprocess control. QB Systems is a product brand of A4BEE Sp. z o.o.

Applications
Bioreactors

Software-defined control for a bioreactor — a new QB vessel, an upgrade to one you have, or a retrofit of the existing PLC.

Buffer & media preparation

Automated preparation of growth media and process buffers, so a recipe runs the same way every time without fixed infrastructure.

Automated sampling

Automated sampling from 4–18 sources, aseptic-capable and up to 72 hours unattended. Works with any vendor's bioreactor.

Deployment
Retrofit

Existing equipment keeps running; QB takes over the PLC, or reads from it without touching control.

Scale
Pilot (50–300 L)

Stainless steel, where QB supplies the control software and integration and a certified partner builds the installation.

Digital maturity: today and target

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

Source: A4BEE analysis of public sources
Process automation 45 → 85
Fermentation at NEOXY runs under automated control, but the post-fermentation stages have historically required manual interventions to maintain stability, which is what the 2025 voluntary shutdowns addressed.
Data integration 30 → 80
Clermont-Ferrand R&D and Carling production maintain different data models, and the air gap between the OT network and the IT network prevents cross-site analytics today.
Predictive analytics 25 → 75
Reactive response to bottlenecks has been the operating pattern through 2025; predictive maintenance and feedstock scoring are absent from the current stack.
Cybersecurity maturity 55 → 80
Physical separation provides a baseline, and CEO fraud, phishing and hacking are named risks; the maturity gain comes from a documented control model rather than from further disconnection.
Regulatory compliance automation 35 → 85
CSRD and multi-standard certification reporting is still manual across sites; sensor-level capture is the change that moves the score.
Remote operations capability 20 → 70
Air-gap architecture prevents Lyon and Clermont-Ferrand from reading Carling production data in real time; a one-way gateway is the enabler, not a workaround.

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 February 2026. It reflects A4BEE's own interpretation and opinion, is not affiliated with, endorsed by, or verified with Afyren, 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].