--- title: "Bioprocess modeling software for fewer runs | Novasign" description: "Novasign is bioprocess modeling software for development and scale-up. Bespark*bio reached the same process decisions with 8 runs instead of 22." url: https://novasign.at/en --- # Develop your process on a model. Confirm it in the lab. Make the right process decisions faster, with up to 70% fewer experiments. Novasign Studio guides experimental design, connects your data and builds hybrid models across your entire process chain, from upstream and filtration to chromatography. Predict how changes affect titer, purity and yield before running experiments, then use traceable results to guide scale-up and support regulatory submissions, all in one platform. [Book a demo](https://novasign.at/en/book-a-demo)[Explore Novasign Studio](https://novasign.at/en/platform) For [biopharma](https://novasign.at/en/solutions/biopharma), [CDMOs](https://novasign.at/en/solutions/cdmos) and [industrial biotech](https://novasign.at/en/solutions/industrial-biotech), including precision fermentation and cultured food. - [Reactor](https://novasign.at/en/platform/reactor) - [TFF](https://novasign.at/en/platform/tff) - [Chromatography](https://novasign.at/en/platform/chromatography) Novasign Studio / Fed-Batch Study / Model and Process Simulator Process Data dataset reader runs 1 to 6 0.3s Online Data dataset reader online.csv Offline Data dataset reader offline.csv PAT Data dataset reader pat.csv Join Data merge dataframes run and time Interpolate interpolate columns offline samples Reactor Model reactor model your structure 2.1s Process Simulator process simulator simulate runs 1.2s 6 suggested changesInspectAccept all Approve for use Reactor Model, trained on the new data Approve Ask the AI Assistant ## From broken handovers to one seamless, flexible workflow Every bioprocess project runs through workflows that cross tools, units and teams: from the experimental plan to a decision, from upstream to downstream, and from early development to manufacturing. Each one breaks at its handovers. Novasign Studio turns them into one workflow that you adapt to how your team works. Planning, execution, data, modelling and decisions usually live in separate tools: a planning tool, the control system, spreadsheets or databases to aggregate and inspect the data, a modelling tool and a meeting for the decision. Each tool needs its own expertise, and every handover is done by hand. Switch to see what changes. Experimental planning manual setupdesign to equipment Experiments copy and pastedata importers Spreadsheets and databases aligned by handdata curation Models hard to retracedigital twin Decisions Next experimentsplanned from scratchmodel-based DoE Next experimentsplanned from scratchmodel-based DoE 1. Conventionally, experimental planning reaches the experiments through manual setup. 2. The results reach spreadsheets and databases by copy and paste. 3. The data reaches the models after being aligned by hand. 4. The decision that follows is hard to retrace. ### How Novasign Studio clears the way - From plan to execution Conventionally a design is set up on the equipment by hand, sometimes with the help of DoE software. Together with device vendors, Novasign is building the link that sends a design from Novasign Studio straight to the equipment, without a manual method setup. - From experiments to data In Novasign Studio every step is a node on one canvas. Data importer nodes read files and live process data, and curation nodes merge them into one dataset. In a biosimilar project with Repligen, Novasign Studio combined bioreactor data, offline viable cell density and titer, MAVEN glucose control and ZipChip CE-MS quality data in one workflow. [Merging five data sources](https://novasign.at/en/proof/case-studies/repligen#data-preparation) - From data to model Extensive preprocessing nodes filter, align and validate the data. You build a reliable ETL pipeline for your process development once, and it runs the same way for every campaign that follows. - From model to decision The model behind the digital twin simulates the scenarios you have not run in the lab and ranks them by the objectives you choose, several at once. Novasign Studio is collaborative, so the whole team opens the same model, works on it together and can retrace the decision. [The digital twin](https://novasign.at/en/platform/reactor#simulate) - From decision to the next experiments Model-based DoE proposes the next experiments from the digital twin’s simulations. Tell it how many runs the next round holds, such as 4, 8 or 12 vessels of a parallel bioreactor system, and it recommends that many. Across customer programs this has meant 35–70% fewer experiments. [Model-based DoE](https://novasign.at/en/platform/experiment-design#model-based-doe) ## Backed by customer results and peer-reviewed science - ### Bespark\*bio Experiment reduction and optimization A model trained on 8 runs instead of 22 made the same decisions for optimizing a viral-vector process as one trained on the full campaign. 64% fewer experiments, about €40,000 of raw material and 12 weeks saved. [Read the study](https://novasign.at/en/proof/case-studies/bespark-bio) - ### Bisy Industrial process optimization Two rounds of designed experiments optimized a Komagataella phaffii process for an industrial enzyme. Enzymatic activity increased 2.4-fold. [Read the study](https://novasign.at/en/proof/case-studies/bisy) - ### Arkeon Optimization and scale-up A designed campaign, model-proposed conditions and validation runs, followed by a scale-up prediction. 5.3-fold specific yield at 150 L over the initial process, and a successful scale-up. [Read the study](https://novasign.at/en/proof/case-studies/arkeon) - ### Repligen Matching CQAs for biosimilars A biosimilar process was designed for a target glycan pattern, with Repligen’s process analytics integrated. A design space for glycoform and glycation, mapped from five experiments and a hybrid model. [Read the study](https://novasign.at/en/proof/case-studies/repligen) - ### ECONTI consortium E2E process Hybrid models of four unit operations ran inside the automation platform of a continuous E. coli process and set the chromatography switching of the next cycle. Intensified DoE raised the upstream titer by 60% over the original process conditions, and the process ran end to end with model-based control. [Read the study](https://novasign.at/en/proof/case-studies/econti) > From the beginning, bespark\*bio has worked closely with Novasign… This transition is powered by Novasign Studio, which we use as our exclusive modeling environment. Matthias Müllner, CEO, Bespark\*bio - [All case studies](https://novasign.at/en/proof) - [Peer-reviewed publications](https://novasign.at/en/proof/publications) ### Why teams choose Novasign - [The next round sized to your vesselsExperiment design](https://novasign.at/en/platform/experiment-design) - [A method published in peer-reviewed journalsPublications](https://novasign.at/en/proof/publications) - [Named customer resultsCase studies](https://novasign.at/en/proof) - [A team that can run the experiments for youProcess development](https://novasign.at/en/services/process-development) Novasign was founded in August 2019 as a BOKU University spin-off, based in Vienna, Austria. It makes Novasign Studio, browser-based bioprocess modelling software. [About Novasign](https://novasign.at/en/company) ## Start with an idea, and let the AI Assistant build it with you Describe what you want to find out, and the optional AI Assistant sets up the workflow with you, from your data to a digital twin. You reach a first model faster, without knowing every node, and every suggestion stays yours to check and accept. Novasign Studio / Fed-Batch Study / Model and Process Simulator Online Data dataset reader online.csv VCD Offline dataset reader vcd.csv Titer Offline dataset reader titer.csv Join Data merge dataframes run and time Units convert column units temp to °C Reactor Model reactor VCD, titer Model Selector model selector best fit Process Simulator reactor process simulator temp, pH Results chart viewer 3D surface AI Suggested Changes InspectAccept All [The AI Assistant](https://novasign.at/en/platform/ai-assistant) ## Built for your industry - [CDMOsDeliver faster, well-justified process decisionsTurn process data, results and expertise into workflows your team reuses for every client.Fewer experiments per client programWorkflows that carry over to the next projectDecisions your sponsors can follow step by stepNovasign Studio for CDMOs](https://novasign.at/en/solutions/cdmos) - [BiopharmaStop running experiments you have already runMake the knowledge from earlier campaigns and platform molecules usable for the next decision.Reuse historical data across molecules and programsBring your existing models, scripts and codeUpstream and downstream in one connected workflowNovasign Studio for biopharma](https://novasign.at/en/solutions/biopharma) - [Industrial biotechReach viable unit economics fasterFind the process variables that move the cost per gram, and scale up with confidence.See which conditions drive production costRecommendations your team can inspect and challengeScale-up predictions before the expensive runNovasign Studio for industrial biotech](https://novasign.at/en/solutions/industrial-biotech) ## Why teams like working in Novasign Studio One workflow, from the data to the model, built in the browser and shared by the whole team. Novasign Studio / Fed-Batch Characterization / Model-Based Design of Experiments Fed-Batch Runs dataset reader runs.csv 0.3s Reactor Model reactor model VCD, titer 2.1s Model Selector model selector best checkpoint 0.5s Process Simulator reactor process simulator temp, pH 1.6s Next Experiments model-based doe 4 runs 0.8s - Work together Colleagues edit the same workflow at the same time, and a node someone is working on is locked for everyone else. - Nothing to install Collaborators open the project in their browser and look at the same data, model and result. - Every result traceable Every change carries its author and time, and every training run is recorded with its data and parameters. - No code, or your own code Build a workflow by dragging nodes, or write your own Python code in the same workflow. [Why teams like working in Novasign Studio](https://novasign.at/en/platform#why-teams-like-it) ## Hybrid models learn from less data and respect the mechanistics A hybrid model combines a given mechanistic structure (mass balances, stoichiometry) with a neural network that learns only the reaction kinetics. Because the mechanistics you know are already in the model, it has less left to learn, so the data from an affordable campaign is enough to train it, and it can often predict beyond the conditions it was trained on. Hybrid Model for a Bioreactor - Inside the Mass Balance. c t+1 V t+1 = c t V t + Σ c in V in − c t Σ V out - Process Parameters. Temperature, pH - Feed. Glucose - Samples. Volume removed - Mass Balance. Given · mechanistic - Machine Learning. Learned · from data. Rates of change - Kinetics. Given · mechanistic. New states - Predicted. VCD. Titer. Glucose [How hybrid modelling works](https://novasign.at/en/platform/hybrid-modeling) ## Every experiment improves the model Each experiment adds to the model. Novasign Studio proposes the runs to perform, the scientist can adapt them, and the results are used to retrain the model. Fewer experiments are needed because the knowledge of the last campaign is carried over. State the goal Design the experiments Run them Prepare the data Train the model Simulate the rest Novasign Studio proposes, the scientist adapts The model proposes the runs to perform and shows how confident it is. Your team can adapt them before they go to the bioreactor. 8 runs instead of 22 to optimize a viral-vector process (Bespark\*bio) 1. State the goal, for example the titer, a quality attribute or the operating window the campaign has to reach. 2. Design the experiments from what is already known. 3. Run the experiments. 4. Prepare the data. Alignment, units and derived quantities are nodes in the workflow, not a script on one person’s laptop. 5. Train the model on the campaign. A person starts every training run. 6. Simulate the runs that were not carried out and read a candidate optimum from the result. A scientist decides which experiments to run next, their results join the dataset, and the model is retrained. Later rounds often need only a confirming run rather than another campaign. - [Experiment design](https://novasign.at/en/platform/experiment-design) - [Let Novasign design it](https://novasign.at/en/services/process-development) ## Start with one process question. See what your data can do. The right starting point if you want concrete results. We work with you on a scoped proof of concept using your own process data, so the outcome is relevant to your process from the first step. 1. 1 ### Scoping call We sign an NDA with you. Then you define one process challenge and one unit operation or workflow to focus on. 2. 2 ### Data review Novasign reviews your existing data and defines what the proof of concept can produce, at no cost up to this point. 3. 3 ### Proof of concept We build and run the workflow in Novasign Studio together with you. 4. 4 ### Results Clear outputs, a defined scope, and your decision on how to proceed. 5. 5 ### Handover The project is handed over to you in Novasign Studio, so your team continues from where the proof of concept ended. [Proof-of-concept modelling](https://novasign.at/en/services/proof-of-concept) ## Studio runs where your data is Novasign Studio runs on your own servers, in your own cloud account or on an appliance that Novasign builds and ships, and the on-premise and appliance options can run air-gapped. You decide where your process data is stored and what leaves your network. - [Deployment and architecture](https://novasign.at/en/platform/deployment-and-architecture) On-premise Your own data centreavailable today Your cloud Your own cloud accountavailable today ## Services that get you to results faster From experiments with our partners to a proof of concept in Novasign Studio and training for your team, our services meet you where your project stands. - [Process developmentNo capacity for experiments, or no data at hand? With our partners in Vienna we carry out the experimental work, planned with you so the data is fit for modelling.Process development](https://novasign.at/en/services/process-development) - [Proof-of-concept modellingOur engineers build the first model of your process with you in Novasign Studio, so your evaluation is faster and the results are yours to continue from.Proof-of-concept modelling](https://novasign.at/en/services/proof-of-concept) - [Novasign Studio trainingRemote or on site, on your own data after a proof of concept or on Novasign’s upstream and downstream datasets without one.Novasign Studio training](https://novasign.at/en/services/training) - [Process modelling workshopA hands-on course at our offices in Vienna, on real data from upstream through filtration to downstream. The next one is planned for the end of Q1 2027.Process modelling workshop](https://novasign.at/en/services/workshop) - [Scientific modellingA new or intensified process, an unusual combination of unit operations, or an idea you want to evaluate before you build it. Our bioprocess engineers design and model the unit with you.Scientific modelling](https://novasign.at/en/services/scientific-modelling) [All services](https://novasign.at/en/services) - [Frequently asked questions](https://novasign.at/en/faq) ## From the blog Articles on methods and results, written by the people who build the models. [All posts](https://novasign.at/en/blog) - [Quality by DesignDigital TwinAug 19, 2026Bioprocess Modeling Regulators Can Follow: Webinar RecapHighlights from our second webinar session: how traceable modeling workflows support experiment reduction, biosimilar CQA targeting, scale-up prediction and continuous processing.](https://novasign.at/en/blog/bioprocess-modeling-regulators-can-follow-webinar-recap) - [BioprocessBioprocess DevelopmentMar 20, 2026De‑Risking Bioprocess Scale‑Up: Predictive Modeling for Reliable Fermentation and Cell Culture ManufacturingScaling a bioprocess from laboratory development to industrial manufacturing is one of the most critical, and risky steps in biotechnology. For contract development and manufacturing organizations (CDMOs), large pharmaceutical companies, industrial biotechnology producers, and emerging sectors such…](https://novasign.at/en/blog/derisking-bioprocess-scaleup-predictive-modeling-for-reliable-fermentation-and-cell-culture-manufacturing) - [BioprocessBioprocess DevelopmentMar 20, 2026Autonomous Continuous Biomanufacturing: Closing the Gap Between Process Development and ProductionContinuous biomanufacturing has long been considered one of the most promising approaches to increase productivity, reduce facility footprint, and enable more sustainable bioprocessing.](https://novasign.at/en/blog/autonomous-continuous-biomanufacturing-closing-the-gap-between-process-development-and-production) ## See how Novasign fits your process and your data 30 minutes on your own campaign data, no qualification step. [Book a demo](https://novasign.at/en/book-a-demo)[Explore Novasign Studio](https://novasign.at/en/platform)