☰

Enantis s.r.o.

Industrial-scale protein engineering, ready for clinical review

Industry
Biotechnology
Headquarters
Brno, Czech Republic
Public information as of
February 2026

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

Strategic priorities

Enantis s.r.o. is the first biotechnology spin-off from Masaryk University in Brno, founded in 2006 and embedded in the INBIT (Incubator of Biotechnology) on the university campus. The company develops engineered proteins using a computational platform rooted in the Loschmidt Laboratories' FireProt, CaverDock and HotSpot Wizard tools, and sells stable growth factors (FGF2-STAB and FGF10-STAB) into stem-cell, cultured-meat and cosmetic markets.

The strategic pivot is from ISO 9001 research-grade manufacturing toward cGMP-compliant pharmaceutical production, anchored by FGF10-STAB for acute respiratory distress syndrome (ARDS) progressing toward First-in-Human trials. The CEO, appointed in October 2024, has framed the therapeutic programme as the company's next commercial step. Existing distribution runs through partners including Qkine (UK) and Bio-Techne.

The external validation is already arriving: the May 2024 Estée Lauder BRKTHRU VOICES Award placed Enantis inside a Tier 1 beauty multinational's vendor-qualification regime. Around €1 million of Horizon 2020 SME Instrument Phase 2 funding supported the preclinical work on FGF2-STAB. Production is centralised in microbial fermentation (E. coli expression), and the company's published work describes wild-type FGF2's half-life of roughly 10 hours at 37 °C against FGF2-STAB's more than 20 days.

What links the three product directions (therapeutic, cultured meat, cosmetic) is the same underlying capability: design data, bioreactor data and quality data that can travel between systems without re-keying. That is the digital gap the page describes, and where the work below focuses.

Challenges we see

  • Operations Manufacturing

    Scaling fermentation while keeping yield under control

    Moving from laboratory-scale fermentation (1–10 L bioreactors) to industrial scale (500–5,000 L) introduces non-linear behaviour: heat transfer, oxygen mass transfer (kLa) and mixing shear stress all change, and E. coli inclusion-body formation is the documented failure mode when parameters drift.

    Where fermentation control depends on individual operators adjusting conditions by hand, the population of conditions a process can credibly cover narrows with each person who leaves. PAT (Process Analytical Technology) instrumentation narrows it back by capturing what the operator used to feel.

  • Digital Integration

    Linking design data to bioreactor performance

    In silico design data sits in Loschmidt servers or FireProt instances while wet-lab validation data sits in separate ELNs (Electronic Laboratory Notebooks) or paper notebooks. The two domains do not currently share a common record.

    When a batch underperforms, tracing the outcome back to the design parameters that predicted stability is a reconstruction exercise rather than a query. A backbone that links the two turns every batch into input for the next round of design.

  • Compliance Regulatory

    Building the data integrity that clinical submissions need

    The therapeutic pivot to FGF10-STAB for ARDS requires moving from ISO 9001 (which certifies that a process exists) to cGMP (which requires validated, continuously controlled processes) with ALCOA+ data integrity (data that is Attributable, Legible, Concurrent, Original and Accurate, plus Complete, Enduring and Available).

    Where records are written by hand on paper, the evidence behind a clinical submission has to be assembled rather than read. Digital records with audit trails produce the evidence as a by-product of running the work.

  • Digital Operations

    Moving proprietary tooling onto enterprise infrastructure

    The protein engineering platform relies on FireProt, CaverDock and HotSpot Wizard, which were developed in an academic environment and run on university infrastructure that does not have the controls a Tier 1 commercial partner's vendor qualification expects.

    Where partner-specific data sits on shared academic systems, each engagement is an open question for the partner's compliance team. Wrapping the same tools in private, role-isolated infrastructure is what makes a commercial relationship answerable on paper.

  • Operations Manufacturing

    Planning across research, cosmetic and pharmaceutical grades

    The portfolio spans research-grade, cosmetic-grade and future pharmaceutical-grade products. The same biological molecule is sold as several SKUs (Stock Keeping Units) with different labels, packaging and quality documentation, distributed through partners such as Qkine and Bio-Techne.

    Planning across multiple grades of the same molecule multiplies the number of stock states to track, and tracking them by spreadsheet grows linearly with the catalogue. An integrated view of grade, lot and demand makes the planning feasible at the rate the business is adding SKUs.

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 fermentation conditions off the bioreactor instead of off the operator

    Enantis's published description of its process refers to 'highly qualified personnel' and manual control, which constrains scale and lifts unit cost.

    Instrumenting reactors with Raman and capacitance probes and streaming the data into a digital twin lets the process adjust feed rates and induction timing against a model rather than against a person's judgement, with a documented envelope for what each run actually delivered.

    • Enantis technical documentation, FGF2-STAB research and process notes
    • Masaryk University INBIT facility description

  2. Connecting design tools and bioreactor data into one R&D backbone

    FireProt predictions and the wet-lab results that follow them currently live in separate places, so each batch's outcome is not automatically connected back to the design that produced it.

    A cloud-native data layer that takes simulation output from the academic tools by API and wet-lab output by IoT (Internet of Things) into one queryable model lets manufacturing data continuously inform the next round of in silico design.

    • Loschmidt Laboratories FireProt, Caver, HotSpot Wizard documentation
    • Enantis_DeepResearch, R&D data backbone section

  3. Producing the records a cGMP submission reads from

    ALCOA+ data integrity is a prerequisite for an EMA (European Medicines Agency) or FDA (US Food and Drug Administration) clinical trial application, and a small team managing paper documents cannot produce that evidence at the volume a cGMP environment generates.

    A GAMP5-validated (Good Automated Manufacturing Practice, 5th edition) electronic QMS (Quality Management System) and LIMS (Laboratory Information Management System) to 21 CFR Part 11 (the US rule on electronic records and signatures) gives the quality organisation a system that already produces the audit trail it has to defend.

    • EMA and FDA guidance on cGMP data integrity, ALCOA+
    • Enantis FGF10-STAB ARDS preclinical programme statements

  4. Putting the protein engineering platform behind enterprise controls

    FireProt and related tools run on university infrastructure that does not have the role-based access control or encryption a Tier 1 commercial partner's vendor qualification expects.

    The same algorithms, deployed on a private cloud tenancy with role isolation, encryption in transit and at rest, and audit logging, give partners the security posture they need to engage commercially without re-engineering the underlying science.

    • Estée Lauder BRKTHRU VOICES Award, May 2024
    • Loschmidt Laboratories academic hosting description

  5. Building brand visibility inside a white-label distribution model

    Distributors including Qkine and Bio-Techne resell FGF2-STAB under their own brand, so Enantis has no direct read on how the product performs in end-user hands or where demand is moving.

    A customer portal and a distributor-data dashboard give end-users a place to read batch-level data while giving Enantis visibility into where the molecule is actually moving, without disturbing the existing distribution agreements.

    • Former CEO on distributor visibility
    • Estée Lauder BRKTHRU VOICES Award, May 2024

What we'd propose

  • Digital Lab

    Digital twin for bioreactor optimisation

    We instrument fermentation reactors with PAT-grade sensors, stream the readings into a time-series model, and run deviation detection and feed-rate optimisation against the batch currently running, so yield and COGS (Cost of Goods Sold) improve with each scale-up rather than regressing.

    • Real-time sensor integration

      Getting data off the bioreactor

      Connect Raman spectroscopy and capacitance probes to bioreactors using OPC UA (Open Platform Communications Unified Architecture) so critical process parameters — dissolved oxygen, pH, cell density — leave the equipment in a documented, vendor-neutral form rather than staying inside a closed controller.

    • AI-driven feed and induction control

      A model that adjusts the run

      Build a model that uses real-time sensor input to recommend feed-rate and induction-timing changes, with a documented envelope of acceptable ranges, so operators see a recommended adjustment rather than guess one.

    • Golden-batch comparison

      Benchmarking every run

      Overlay the current batch's trajectory against the best historical runs so deviation from the best-known profile surfaces within the run rather than after the harvest.

    • Yield and COGS become a property of the process design rather than a property of the operator on shift.
    • Each scale-up step inherits the data and the model from the previous step, so 10 L to 500 L to 5,000 L is a sequence rather than a fresh project.
    • Run records arrive as data, which is what the next opportunity (GMP data integrity) also needs.

    We've done this before

    See how we do Digital Lab
  • Digital Lab

    GMP-ready data infrastructure

    We deploy a GAMP5-validated electronic QMS and LIMS that satisfy 21 CFR Part 11 and EU Annex 11 (the EU equivalent for electronic records and signatures in regulated environments), so the records behind a clinical trial application are produced by the work itself rather than assembled for it.

    • ALCOA+ data integrity

      Records that hold up

      Implement electronic record capture so every entry is Attributable, Legible, Concurrent, Original and Accurate, with tamper-evident audit trails and electronic signatures that meet EMA and FDA expectations.

    • Automated CAPA and document control

      Compliance work as a workflow

      Configure the QMS so deviations, change controls and CAPAs (Corrective and Preventive Actions) follow a documented workflow with named approvers and a complete audit trail, so a non-conformance routes itself rather than relying on memory.

    • Certificate of Analysis automation

      CoA from data, not from re-keying

      Generate CoAs (Certificates of Analysis) from the LIMS data they describe, with the release workflow tied to the analytical result rather than to the paperwork that follows it.

    • Records behind a clinical submission are produced continuously rather than compiled under deadline pressure.
    • Audit preparation becomes a query against the system instead of a reconstruction exercise.
    • The same record set serves QA, regulatory and the cGMP submission, with one chain of evidence.

    We've done this before

    See how we do Digital Lab
  • Enterprise AI

    R&D data backbone integration

    We build a cloud-native backbone that connects the Loschmidt computational tools (FireProt, CaverDock, HotSpot Wizard) to the wet-lab instruments that validate them, so every batch feeds the next round of design through a query rather than through a person.

    • Computational tool API integration

      FireProt into the backbone

      Define and implement APIs (Application Programming Interfaces) against the academic tools so stability predictions, tunnel analyses and mutagenesis suggestions land in the backbone as structured data with their provenance attached.

    • Wet-lab instrument connectivity

      Instruments into the backbone

      Connect the analytical instruments (NMR, mass spectrometry, HPLC) and the bioreactors via IoT so validation data enters the backbone as data rather than as printed reports.

    • Closed-loop ML feedback

      Design learns from production

      Use the joined dataset to surface correlations between predicted stability and observed performance, with each new batch updating the next round of design recommendations.

    • Design-to-production traceability moves from a person remembering to a system recording.
    • Past simulations become queryable training data for the next generation of models.
    • Root-cause analysis of a failed batch starts from the design parameters, not from a reconstruction.

    We've done this before

    See how we do Enterprise AI
  • Enterprise AI

    Secure cloud migration for the protein engineering platform

    We migrate FireProt, CaverDock and HotSpot Wizard from academic infrastructure onto a private cloud tenancy with enterprise controls, so the same scientific capabilities are delivered inside the security posture a Tier 1 commercial partner's vendor qualification requires.

    • Private cloud deployment

      The same tools, behind enterprise controls

      Stand up a dedicated cloud tenancy with encryption in transit and at rest, network isolation, and logging, so partner-specific projects run in an environment that satisfies vendor-qualification review.

    • Role-based access control

      Per-partner isolation

      Implement RBAC (Role-Based Access Control) so a partner can only see the project and data that belongs to them, with access events written to an audit log a partner's compliance team can read.

    • Elastic compute for simulations

      HPC on demand

      Provide elastic high-performance computing capacity so computationally heavy FireProt stability predictions run at the throughput a multi-partner pipeline needs, without the academic cluster's queue.

    • Partner-specific projects satisfy commercial security review without changing the underlying science.
    • Compute scales with the project pipeline instead of competing with academic scheduling.
    • The IP (intellectual property) boundary is enforced by infrastructure, not by organisational convention.

    We've done this before

    See how we do Enterprise AI
  • Agents

    AI agents for cGMP document work

    We deploy narrow, reviewable agents that take the repetitive part of the cGMP document work: drafting deviation and change-control summaries from source records, checking a document against its template before review, and finding every controlled document a standards change touches. A named person approves every output.

    • Drafting from source records

      First drafts from system data

      Generate the first draft of a deviation report, change control, periodic review or CoA summary directly from the underlying records in the LIMS and QMS, so the author edits and judges rather than assembles.

    • Template and completeness checking

      Gaps found before review

      Check a submitted document against its template and the site's checklist before it enters the human review queue, returning missing or inconsistent sections so they are addressed by the author rather than by the reviewer.

    • Change impact search across the document set

      Which documents a change touches

      When a standard, method or specification changes, retrieve every controlled document that references it and rank them by how directly they are affected, so the update scope is established on day one rather than discovered in audit.

    • Review queues move faster because documents arrive complete.
    • The scope of a standards change is established by search rather than by recollection.
    • 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

Digital maturity: today and target

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

Source: A4BEE analysis of public sources
Process automation 30 → 75
Enantis's published description of its fermentation process refers to 'highly qualified personnel', consistent with manual control on laboratory-scale reactors. The PAT-instrumented, model-driven target replaces that dependency with documented inputs and a model.
Data integration 25 → 80
In silico design output and wet-lab validation output sit in different domains, with no Golden Thread linking a predicted mutation to its observed performance. The target backbone makes the link a query rather than a reconstruction.
Regulatory compliance 40 → 90
The site is ISO 9001 certified, which addresses process consistency rather than cGMP validation and ALCOA+ evidence. The target is GAMP5-validated LIMS and QMS to 21 CFR Part 11, which is what a clinical submission reads from.
Cybersecurity 35 → 85
Protein engineering tools run on university infrastructure without commercial-grade access controls. The target is a private cloud deployment with RBAC, encryption and partner-isolated projects.
Supply chain visibility 30 → 70
The white-label distributor model means Enantis has no direct read on end-user demand. The target is an integrated distributor data feed plus a customer portal that exposes batch data to end-users.
Scalability infrastructure 35 → 80
Pilot-scale operations constrain both the volume available for cultured-meat customers and the unit cost a clinical programme can absorb. The target is industrial-scale production with a digital twin that makes scale-up a sequence rather than a fresh project.

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 Enantis s.r.o., 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].