☰

AMSilk GmbH

Scaling silk protein production through partners

Industry
Industrial biotechnology (precision fermentation)
Headquarters
Neuried (Munich), Germany
Public information as of
January 2026

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

Strategic priorities

AMSilk raised EUR 52 million in September 2025 to move from pilot volumes to industrial supply of recombinant silk proteins. The company has said the money is for commercial growth and industrial scale-up, and the expanded Evonik agreement announced the same month targets multi-ton monthly output from the fermentation line in Slovenská Ľupča, Slovakia.

Production runs at partner sites rather than at AMSilk's own plants. Precision fermentation and downstream processing sit with Evonik in Slovakia and, since the October 2024 agreement, with Ajinomoto in France, while strain engineering, process development and application testing stay at the Neuried campus near Munich.

The immediate operational priority follows directly from that split: AMSilk needs process and quality data from partner sites, and from its own development labs, in a form it can act on while a batch is still running rather than after it has shipped.

Challenges we see

  • Operations Manufacturing

    Getting fermentation data from partner sites during a run

    Fermentation and downstream processing happen inside Evonik's and Ajinomoto's control systems, while the people who designed the process work in Neuried. Batch performance currently reaches Munich through reports issued after the run.

    As monthly output rises under the expanded Evonik agreement, more commercial commitments will depend on information that arrives after a batch has finished, unless a live data path is agreed with each manufacturing partner.

  • Digital Integration

    Lab instruments are not yet feeding the electronic lab notebook

    AMSilk has moved laboratory records onto the Scifeon electronic lab notebook and lab information system, replacing spreadsheets and paper notebooks. Analytical instruments such as chromatography systems, mass spectrometers and tensile testers still hand over their results manually.

    The return on the new lab system scales with how much data arrives in it automatically, so instrument connectivity becomes the next step rather than a separate project.

  • Data Integration

    No single batch record spanning AMSilk and its manufacturing partners

    Raw material lots, fermentation runs, purification steps, spinning and final fibre testing are recorded in different systems owned by different organisations. Linking a finished yarn back to the conditions of the run that produced it is a manual exercise.

    Questions about yield and quality across sites take longer to answer than the production cycle allows once several sites are running the same process in parallel.

  • Compliance Regulatory

    Per-batch environmental data for the EU Digital Product Passport

    The EU Ecodesign for Sustainable Products Regulation introduces a Digital Product Passport for textiles, carrying documented circularity and environmental footprint data. AMSilk's brand customers will pass that requirement up the chain.

    Environmental figures will need to be produced per batch from recorded production data rather than from periodic studies, which changes what has to be captured at each partner site.

  • Operations Manufacturing

    Transferring a process to a second manufacturing site

    Bringing the Ajinomoto site in France onto the same process means moving recipes, critical process parameters and standard operating procedures from Munich to equipment with different geometry and different control systems.

    Each additional site repeats the transfer effort as long as the process definition lives in documents rather than in a machine-readable recipe format.

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. Seeing partner fermentation data while a batch is running

    Batch performance at Evonik and Ajinomoto reaches Munich through reports issued after the run, so deviations in pH, dissolved oxygen, temperature or agitation are reviewed once the batch is already complete.

    A secure, read-only data path from partner control systems into a monitoring view in Munich lets AMSilk compare a running batch against reference runs and raise a question with the site the same day.

    • AMSilk press release, 'Evonik and AMSilk extend partnership for sustainable biotech silk materials', September 2025
    • AMSilk press release, 'AMSilk Secures EUR 52M Strategic Financing to Accelerate Commercial Growth', September 2025

  2. Getting instrument results into the lab notebook without retyping

    The Scifeon electronic lab notebook is in place, but analytical instruments in Neuried still deliver results that someone transfers by hand, which costs scientist time and leaves gaps in the audit trail.

    Direct connections from instruments and bioreactors into the lab notebook remove the retyping step and give process development a queryable record of every run.

    • Scifeon customer story on the AMSilk electronic lab notebook implementation, 2024

  3. One batch record covering AMSilk and both manufacturing partners

    Material lots, fermentation runs, purification, spinning and fibre testing sit in separate systems across separate organisations, with no common way to describe a batch.

    A shared data model and a central store give AMSilk end-to-end genealogy for every kilogram of protein, and a base for yield and quality analysis across sites.

    • AMSilk press release, 'AMSilk announces Successful Partnership Negotiation with Ajinomoto for Industrial Fermentation', October 2024

  4. Producing environmental and quality documentation per batch

    Partner sites supply batch, energy and quality information as documents. Turning those into per-batch environmental figures and customer-facing quality records is manual work that grows with every additional batch and customer.

    Agents that read partner documents, extract the figures into the batch record, check them against the data points the Digital Product Passport requires and draft the customer response turn a growing manual load into a review task.

    • EU Ecodesign for Sustainable Products Regulation, Digital Product Passport requirements for textiles

  5. Bringing a new manufacturing site onto the process faster

    Process transfer to a new site relies on documents and email, and equipment differences between sites have to be resolved case by case during ramp-up.

    A machine-readable process recipe plus structured transfer reporting shortens ramp-up at each new site and makes the comparison between sites explicit rather than anecdotal.

    • AMSilk press release on the Ajinomoto industrial fermentation partnership, October 2024

What we'd propose

  • Digital CDMO

    Remote monitoring of fermentation at partner sites

    A secure, read-only data path from Evonik's and Ajinomoto's control systems into a monitoring view in Munich, so AMSilk can follow a batch as it runs without operating the plant.

    • Secure connectivity to partner control systems

      Read-only link, agreed scope

      An edge connection at the partner site publishes an agreed set of process tags outward only, with no inbound control path. Scope, retention and access are fixed in the data agreement with each partner.

    • Batch monitoring view

      Live run against reference runs

      Fermentation parameters such as pH, dissolved oxygen, temperature and agitation are shown against reference batches, so a divergence is visible while the run is still in progress.

    • Deviation alerts routed to Munich

      Notify the process owner directly

      Threshold and trend rules notify the AMSilk process development team when a parameter leaves its expected band, with the batch context needed to open a conversation with the site.

    • Process questions are raised during the run instead of after the batch report arrives.
    • Root-cause work on a yield deviation starts from recorded process data rather than from reconstructed timelines.
    • Production visibility scales with output without AMSilk building or staffing its own plant.

    We've done this before

    See how we do Digital CDMO
  • Digital Lab

    Connecting lab instruments and bioreactors to the Scifeon lab notebook

    Integration work that takes results from analytical instruments and development bioreactors in Neuried directly into the electronic lab notebook, removing manual transfer.

    • Instrument connectors

      One driver per instrument type

      Connectors for chromatography systems, mass spectrometers, tensile testers and development bioreactors, using each instrument's available interface rather than a single assumed standard.

    • Automatic capture into the lab notebook

      Result lands with its context

      Each result is written into Scifeon against the experiment, sample and instrument that produced it, so the record is complete without a scientist retyping values.

    • Data integrity trail

      Attributable, timestamped, unedited

      Captured values keep their origin, timestamp and any instrument-side revisions, which is the record structure regulated customers and audits expect.

    • Scientist time moves from data entry to experiment design.
    • Historical development runs become queryable rather than sitting in exported files.
    • The lab notebook investment starts returning data at the rate the lab generates it.

    We've done this before

    See how we do Digital Lab
  • Enterprise AI

    A shared data model across AMSilk and its manufacturing partners

    An ontology-based platform that harmonises material, process and quality data from AMSilk, Evonik, Ajinomoto and the spinning partners into one queryable record per batch.

    • Common definitions for batches and materials

      One vocabulary across partners

      Site-specific tag names and file formats are mapped onto shared classes for material lot, run, process step and quality result, so a query works the same way against every source.

    • Batch genealogy

      Trace fibre back to feedstock

      Every finished quantity links back through spinning, purification and fermentation to the raw material lots and process conditions behind it.

    • Analysis-ready data layer

      Clean inputs for yield work

      Harmonised, contextualised data feeds yield and quality analysis across sites, including comparison of the same process run on different equipment.

    • A quality question can be answered from one place instead of three sets of exports.
    • Yield differences between sites are measurable rather than inferred.
    • New manufacturing partners connect to an existing model instead of adding another silo.

    We've done this before

    See how we do Enterprise AI
  • Agents

    Agents for partner batch documents and product documentation

    AI agents that read the batch, energy and quality documents partner sites supply, extract the figures into the batch record, and draft the environmental and quality documentation customers request.

    • Document extraction into the batch record

      PDF figures become structured data

      Agents pull batch numbers, process summaries, energy figures and quality results out of partner documents and write them into the batch record, with the source page kept for verification.

    • Completeness check against passport requirements

      Flag what is still missing

      Each batch is checked against the data points the Digital Product Passport and customer questionnaires require, and gaps are listed by site and batch before the customer asks.

    • Drafted customer documentation with human sign-off

      Draft first, person approves

      Environmental and quality responses are drafted from the verified batch record and routed to a named reviewer, who approves or corrects before anything reaches the customer.

    • Customer documentation keeps pace with batch volume without adding headcount.
    • Missing environmental data points surface while the batch is recent enough to chase.
    • Every figure in a customer document traces back to the partner document it came from.

    We've done this before

    See how we do Agents
  • Digital CDMO

    Digital process transfer to new manufacturing sites

    Turning AMSilk's production process into a machine-readable recipe plus structured transfer reporting, so bringing a new partner site onto the process is a repeatable exercise.

    • Machine-readable process recipe

      Recipe separate from equipment

      Critical process parameters and process steps are captured in an equipment-independent recipe structure, then mapped to each site's vessel geometry and control system.

    • Structured transfer reporting

      Compare runs across sites

      Runs at the receiving site are reported against the sending site's reference in a fixed structure, so ramp-up progress and remaining gaps are visible to both organisations.

    • Remote guidance for site operators

      Munich expertise, no travel

      Assisted-reality work instructions and live remote support let Neuried process specialists walk partner operators through a procedure without being on site.

    • Ramp-up at a new site starts from a defined recipe rather than a document pack.
    • Differences between site equipment are handled in the mapping instead of during production.
    • Process knowledge stays with the recipe when the team changes.

    We've done this before

    See how we do Digital CDMO

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.

Filtration (TFF)

Benchtop tangential flow filtration for concentration, diafiltration and buffer exchange.

Deployment
Retrofit

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

Scale
Benchtop (1–8 L)

Glass vessels with the complete hardware and software stack. This is the core range for development work.

Digital maturity: today and target

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

Source: A4BEE analysis of public sources
Lab instrument data capture 45 → 85
The Scifeon electronic lab notebook is in place, replacing spreadsheets and paper, but analytical instruments still hand over results manually.
Visibility into partner production 25 → 80
Fermentation runs at Evonik and Ajinomoto; process data reaches Munich through reports issued after the run rather than during it.
Shared batch data model 20 → 75
Material, process and quality data sit in separate systems at AMSilk, the fermentation partners and the spinning partners, with no common batch definition.
Supply and inventory planning 32 → 72
A multi-node chain from fermentation to spinning to brand delivery, with biological run-to-run variability, is being coordinated against fashion and automotive delivery dates.
Per-batch environmental records 35 → 85
Sustainability positioning is central to the product, while the EU Digital Product Passport moves the evidence requirement from periodic studies to per-batch data.
Process transfer recipes 40 → 80
Process definitions travel to partner sites as documents; a machine-readable recipe format would make each additional site transfer repeatable.

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