☰

Bio-Gen Sp. z o.o.

Scaling to 30 markets, one recipe

A Polish agricultural biotech growing from a single lab to 30 markets while keeping its process consistent

Industry
Agricultural Biotechnology
Headquarters
Łódź, Poland
Public information as of
February 2026

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

Strategic priorities

Bio-Gen Sp. z o.o. has run its bioprocess lab and factory in Łódź since relocating there in 2018, and now sells biofertilizers and biostimulants in more than 30 countries. In 2023 the company posted 47.6 million PLN of operating revenue, 12.0 million PLN of EBITDA and 7.3 million PLN of net profit, with year-on-year revenue growth averaging around 7.2 million PLN. The business is self-funded at roughly 23 percent debt to assets, which leaves room for capital programmes alongside operating spend.

The 35-year-old company is now scaling two programmes at once. Bio-Lider, the field-advisory subsidiary in which Robert Lewandowski is a strategic investor, is on track to reach 101 advisors by the end of 2025. Industrial data work is being lifted from isolated bioreactor screens into a platform that can carry process data, laboratory results and field trial data on the same ontology. The two programmes converge on the same question: how to keep the recipe consistent when the company is producing for climates as different as Poland, Saudi Arabia and Brazil.

The strategic frame is explicit. Bio-Gen describes its trajectory as a move from traditional biotechnology to a TechBio model, in which AI, machine learning and real-time analytics become part of the production process itself rather than tools bolted on afterwards. That framing puts the industrial data platform, the laboratory digitisation and the fermentation simulation work on the same critical path.

Challenges we see

  • Operations Manufacturing

    Keeping fermentation reproducible across 24/7 bioreactor operations

    Bio-Gen's bioreactors run continuously to maintain declared CFU (colony-forming unit) concentrations and microbiological purity in bacterial products. Transferring laboratory results to industrial scale while preserving those guarantees requires precise environmental control, and any deviation in temperature, pH, dissolved oxygen or feed timing carries the risk of a contaminated batch.

    Where each parameter is observed through its own interface, the room between a small drift and a batch write-off is wide. Reading the bioreactor as a single time-series against the known-good signature of the run narrows that room and gives the operations team the same view of the process that the engineers have.

  • Digital Operations

    Replacing manual CFU counting with computer vision in the Łódź lab

    The Łódź laboratory handles large daily sample volumes of colony counting (CFU) and morphological strain identification. Highly trained technicians perform repetitive microscopic analyses that consume R&D capacity that could otherwise go into strain development and new formulations.

    Once microscope images become data rather than observation, the analytical step stops being a function of who is on shift. That returns the trained scientists' hours to the work the company hired them for, and keeps the analytical baseline steady across shifts and seasons.

  • Operations Integration

    Turning scattered field-trial data into advisor-ready evidence

    Bio-Lider advisors work with farmers whose average age in Poland is 53 and whose habits favour mineral fertilisers and chemical crop protection. Trial results sit in paper reports and basic spreadsheets across the network, which makes the conversion conversation slower than the advisors would like.

    Where trial results can be retrieved and visualised on the same device the advisor uses at the field edge, the conversation with the farmer moves from description to demonstration, which is what the average farm gate takes to convert.

  • Digital Integration

    Opening OT systems to cloud and remote access without losing the recipe

    Bio-Gen's production infrastructure has accumulated over 35 years, so connectivity across bioreactors, filtration and filling lines varies. Opening operational technology (OT) systems to cloud platforms and remote AR (augmented reality) support increases the attack surface in a sector where bacterial strain IP and process know-how are commercially decisive.

    Connecting legacy equipment through a documented, vendor-neutral protocol is what makes the security conversation about design rather than improvisation, and lets the company decide deliberately what is reachable from outside and what is not.

  • Digital Integration

    Linking feedback from 30+ markets back to the production batch

    Bio-Gen operates in more than 30 countries including Poland, Qatar, Saudi Arabia and Brazil, each with different soils and climates. Field feedback on product performance is not currently correlated with specific production batches, which limits the ability to optimise formulations for a given region or to predict how a new market will respond to an existing product.

    Where market feedback and production records share an ontology, the formulation team can ask a regional question of the data and get an answer that is traceable back to the specific batch and process run that produced the product being evaluated.

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 process signals as a continuous time series

    Bioreactor monitoring currently relies on separate interfaces for each parameter, which forces operators to piece together the state of a run across screens and means a subtle drift can pass unnoticed until it becomes a contamination event.

    An industrial data platform that streams each bioreactor's parameters through OPC UA (Open Platform Communications Unified Architecture) into a shared time-series model lets the team see each run against a known-good signature, flag drift early, and keep the evidence of every batch in one place.

    • Bio-Gen research file, bioprocess reproducibility section, 2026
    • Bio-Gen, About us (bio-gen.pl/en/kim-jestesmy/about-us/), January 2026

  2. Automating colony counting and morphological classification in the Łódź lab

    Manual CFU counting and morphological identification consume highly trained technician time on repetitive work, with accuracy that drifts over long shifts and R&D capacity that goes into counting rather than strain development.

    Computer vision trained on Bio-Gen's own bacterial strains counts colonies and classifies morphology at microscope scale, frees scientists for higher-value work, and keeps the analytical baseline consistent across shifts and seasons.

    • Bio-Gen research file, laboratory analytics section, 2026
    • Bio-Gen, Career page, January 2026

  3. A mobile evidence platform for Bio-Lider advisors in the field

    Bio-Lider advisors carry trial results on paper and in spreadsheets, which limits how quickly and how visually they can demonstrate product performance to farmers who are weighing biologicals against familiar chemical alternatives.

    A tablet and phone platform with regional trial results, product comparisons and simple visualisation gives the advisor the evidence in hand at the field edge, and lets the same data flow back as structured feedback for the formulation team.

    • Bio-Gen research file, Bio-Lider and field advisory section, 2026
    • Top Agrar Polska, 35 lat firmy Bio-Gen, 2025

  4. Designing the new production capacity for reconfiguration from day one

    Current production lines take weeks to reconfigure for a new product, which limits the company's ability to respond to seasonal demand peaks or to emergency needs such as soil regeneration after floods or droughts.

    Planning the next capacity expansion around the Module Type Package (MTP) standard, with documented module interfaces and an orchestration layer, makes new bioreactors and processing units plug in within days rather than months.

    • Bio-Gen research file, modular production section, 2026
    • Bio-Gen 2023 financial data, BizRaport KRS 0000208021

  5. Building a digital twin of the fermentation process

    New culture media and fermentation parameters can only be tested by committing physical bioreactor capacity, which extends R&D cycles and ties up expensive equipment during early-stage work.

    A digital twin of Bio-Gen's fermentation process lets the R&D team run virtual experiments on new formulations, predict process anomalies before they appear on the line, and identify operating conditions that reduce energy and water consumption without changing product quality.

    • Bio-Gen research file, Digital Twin section, 2026
    • Biotechnologia.pl interview with Jarosław Peczka and Ewa Kaniowicz, 2025

What we'd propose

  • Enterprise AI

    Industrial data platform for the Łódź bioreactor hall

    A data platform that brings Bio-Gen's bioreactors, filtration and filling equipment onto OPC UA, streams their parameters into a shared time-series model, and gives the operations and quality teams the same live view of each batch.

    • OPC UA connection layer

      Getting data off the equipment

      Connect bioreactors, filtration skids and filling lines through OPC UA (Open Platform Communications Unified Architecture) so parameter values leave each piece of equipment in a documented, vendor-neutral form, with the legacy PLCs (programmable logic controllers) bridged through the same layer rather than kept on isolated protocols.

    • Golden-batch comparison

      Each run against its known-good signature

      Overlay each live fermentation run against the historical signature of the same product so the team can see drift in temperature, pH, dissolved oxygen and feed rate against a reference curve instead of a static threshold.

    • Anomaly detection on the running batch

      Flags before the run finishes

      Use the same parameter stream to detect contamination signatures and equipment degradation as they emerge, so the response happens against the batch that is still being made rather than the batch that already failed.

    • Deviations surface against the batch that is running, not the batch that already shipped.
    • The same data feeds operations, quality and the formulation team instead of three separate extracts.
    • The evidence behind every batch release is generated by the line rather than compiled for it.

    We've done this before

    See how we do Enterprise AI
  • Digital Lab

    Computer vision and LIMS for the Łódź laboratory

    A laboratory system that automates colony counting and morphological classification, captures results at the instrument, and connects the lab record to the production batch so analytical evidence flows without transcription.

    • Vision-based colony counter

      Counts and classifies from microscope images

      Deploy object detection on Bio-Gen's own bacterial strains so colony counts and morphological classes come from the microscope image directly, with the same model updated as new strains enter the catalogue.

    • LIMS (Laboratory Information Management System) with sample-to-batch link

      Lab result to batch record

      Capture each analytical result with sample identity, method version and timestamp, and link it to the production batch so a release decision can be traced back to the specific run that produced each number.

    • Audit trail for laboratory data

      Electronic records that hold up

      Apply electronic signature, versioning and audit-trail handling to ALCOA+ (the data integrity standard for regulated laboratories) so the laboratory evidence chain stands on its own during an audit.

    • Trained scientist hours move from counting into strain development and formulation work.
    • Analytical accuracy holds steady across shifts and seasons because the baseline is the model, not the operator.
    • Laboratory evidence is queryable by sample, batch and method instead of being reconstructed for each review.

    We've done this before

    See how we do Digital Lab
  • Digital CDMO

    MTP modularity for the next production expansion

    An MTP-based design for Bio-Gen's next capacity step: a documented module interface library, an orchestration layer that recognises new equipment as it is added, and a procurement language that makes vendor neutrality a purchase condition.

    • MTP module library

      Documented interfaces for each equipment type

      Build the MTP (Module Type Package) service descriptions for bioreactors, filtration skids and filling lines so each piece of equipment exposes its control interface in a standard, vendor-neutral form.

    • Plug-and-produce orchestration

      New equipment integrated in days

      Stand up the orchestration layer that discovers a new MTP module on the network and brings it into the production sequence without bespoke programming, so a new bioreactor can be brought on line in days rather than months.

    • Procurement language for interoperability

      Interoperability as a purchase condition

      Write the OPC UA information model and MTP service requirements into the equipment procurement specification, so vendor neutrality arrives with the machine rather than being argued for after delivery.

    • Reconfiguration time drops from weeks to days, which makes seasonal and emergency product switches feasible.
    • The plant runs on a documented interface map rather than on a chain of bespoke integrations.
    • Future capacity additions attach to the same architecture instead of triggering another integration project.

    We've done this before

    See how we do Digital CDMO
  • Digital Lab

    Digital twin of the fermentation process

    A simulation of Bio-Gen's fermentation that runs alongside the physical bioreactors, lets the R&D team test new formulations in software, and gives the operations team early warning of drift and energy waste.

    • Fermentation simulation engine

      Cell growth, nutrients, metabolites in software

      Build a physics-based model of Bio-Gen's fermentation that represents cell growth kinetics, nutrient consumption and metabolite production, validated against historical batch data so the simulation predicts what the next run will do.

    • Virtual experiment platform

      New formulations tested before they hit the reactor

      Run new culture media compositions, temperature profiles and feeding strategies in the digital twin before committing a 1000-litre bioreactor, so the R&D team screens candidates without using production capacity.

    • Energy and water optimisation

      Lower-resource operating points, same product quality

      Use the same model to identify operating points that hold product quality while reducing energy and water consumption, supporting the European Green Deal positioning that runs through the company's commercial story.

    • Failed experiments are caught in software before they use a bioreactor.
    • R&D cycle time shortens because the first physical run is the third or fourth candidate, not the first.
    • Energy and water use is steered rather than reported after the fact.

    We've done this before

    See how we do Digital Lab
  • Agents

    Mobile evidence platform for the Bio-Lider field network

    A tablet and phone platform that gives Bio-Lider advisors structured access to trial results and product comparisons at the field edge, and feeds the same data back as structured feedback for the formulation team.

    • Regional trial results on the advisor's device

      Evidence at the farm gate

      Surface regional trial results, soil and climate context, and product comparisons through a tablet and phone interface so the advisor can show a farmer what has worked on similar land, in the advisor's hand, before the conversation moves on.

    • Structured field feedback into the formulation pipeline

      Field observations become data

      Capture advisor observations and farmer-reported outcomes as structured records rather than free-text notes, and route them into the same data model that holds the production batch, so the formulation team can correlate field performance with specific production runs.

    • Advisory brief generation from the corpus

      Draft briefs the advisor edits

      Draft the regional brief the advisor takes to the farm from the trial corpus and the field feedback already on file, so the advisor spends the preparation time on judgement rather than assembly. A named advisor approves every brief before it leaves the company.

    • Advisors arrive at the farm with the evidence in hand, which changes the conversion conversation from description to demonstration.
    • Field observations are queryable by region, season and product rather than filed away as anecdote.
    • Brief preparation shifts from assembly to judgement, and every output is signed off by a named advisor.

    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.

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 Bio-Gen Sp. z o.o.'s own published ambition implies — not a perfect score.

Source: A4BEE analysis of public sources
Data integration 35 → 85
Bioreactor, laboratory and field-trial data are produced by separate systems without a shared ontology. The industrial data platform work is ahead of the company rather than behind it, and is the precondition for the rest of the roadmap.
Process automation 40 → 80
Manual CFU counting, paper-based trial records and isolated bioreactor screens are the current state. MTP standardisation and laboratory automation together move the floor to one where the next capacity expansion arrives modular.
Analytics and AI 25 → 75
Today, deviation detection is a manual read of individual screens. A digital twin of the fermentation process and anomaly detection on the running batch are the steps that put AI inside the production loop rather than next to it.
Field digitalization 30 → 70
Bio-Lider advisors work from paper reports and spreadsheets in front of farmers whose average age is 53. A mobile platform with structured trial data is what changes the field conversation.
Cybersecurity 45 → 90
Connecting legacy equipment and opening OT to remote access raises the bar for IEC 62443 (the industrial cybersecurity standard) and Zero Trust segmentation, particularly with bacterial strain IP and process know-how on the line.
IT/OT convergence 35 → 80
35 years of accumulated equipment means heterogeneous PLCs and fieldbus protocols. The OPC UA middleware work and MTP standardisation together define the IT/OT boundary by design rather than by accident.

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 Bio-Gen Sp. z o.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].