INDUSTRIAL BIO

AI-assisted protein engineering for industrial teams

Trusted by scientists at Novonesis, Corteva, 
IFF and more.

  • Cortiva Catalyst, a Cortiva Sciences brand logo

Fast-track your protein
engineering with Cradle

Your new AI-assisted therapeutics pipeline

Fast-track your protein
engineering with Cradle

Get protein-based products into the field, factory, or end product in less time using Cradle’s protein design software platform. Generate molecules that meet multiple objectives, powered by your data and expertise.

Protein engineering before

Project design

Target ID

Variant generation

Variant generation

Optimization

Optimization

Application testing

Lead Op

Scale up

Scale up

Protein engineering with Cradle

Project design

AI protein engineering

Scale up

Project design

AI Protein
Engineering

Scale up

Project design

AI protein engineering

Scale up

Goodbye sequential, hello simultaneous

Generate proteins that move the needle on performance and manufacturability, in every design cycle.

Catalytic activity

Catalytic activity

Stability

Stability

Yield

Yield

Efficiently explore diverse designs

Cradle builds on the evolutionary context of your protein and learns each time you upload experimental data. Take bigger strides towards your project goal by exploring and learning from multiple mutations in every tested variant. Added bonus: lower risk of project failure and a stronger IP position.

Optimize multiple objectives at once

Define all the properties that matter for your protein—catalytic activity, stability under harsh conditions, solubility, and beyond. Generate diverse variants that tick all the boxes, starting from just 96 variants per iteration.

Design for real-world function

Your protein’s ultimate test is rarely a microliter assay. What matters is its performance in the fermentor, field, or end product. Whether that means stability in harsh detergent washes, tolerance to solvents, in planta performance, or robustness during industrial fermentations. Design proteins that push towards the properties that matter from day one.

Start training

  • P450_substr_spec_R1.csv

    Metab_path_opt_R2.xlsx

    Biocat_act_screen_R3.csv

    Proc_temp_stab_R1.csv

    Enz_expr_lvls_R2.xlsx

    Chem_conv_rates_R1.csv

    Ind_yield_opt_R2.csv

Train Cradle across your datasets

Connect all your experimental results in one place and train custom AI on your data, from FACS screening to high resolution assays.

  • P450_substr_spec_R1.csv

    Metab_path_opt_R2.xlsx

    Biocat_act_screen_R3.csv

    Proc_temp_stab_R1.csv

    Enz_expr_lvls_R2.xlsx

    Chem_conv_rates_R1.csv

    Ind_yield_opt_R2.csv

Start training

Works with any format and assay

Cradle plays well with your experimental data - no matter the source. Import NGS data from FACS screening campaigns or minimally processed read-outs from chromatography, affinity- or physicochemical assays, and more.

Train AI throughout your program

Connect the dots and maximize learning throughout your program and across your organization. From early discovery stages to application testing, and from positive hits to negative results, every datapoint strengthens your outcomes with Cradle.

No historical data? No problem.

Start a new program from a single sequence, or kick-start your Cradle journey with data you have already collected.

Protein design at any level of expertise

Use Cradle with simple defaults, or set advanced design requirements based on your in-house expertise.

Blocked

Blocked

14 positions

14 positions

Unreliable

Unreliable

16 positions

16 positions

Ignore

Ignore

21 positions

21 positions

Simple workflows

Run Cradle through its user interface, or exercise more control via the API. No need for deep computational expertise. If you want extra support, our team of advisors and engineers is here to guide you.

Mutations where you want them

Allow models to introduce mutations freely or specify where to allow variation, whether in the active site, pre-set domains, or in custom regions defined by you. You can apply precise rules to ensure that proteins are compatible with downstream processing and scale-up.

AI your whole team understands.

Quickly generate clear, actionable reports to analyze your data and understand model performance, giving you the confidence you need before taking sequences into the lab.

Works on any protein and all properties

Working on something not listed? If you can measure it there's a good chance Cradle can help you speed up the path to your program goals. Contact our sales team to discuss your specific needs.

Enzymes

Binders

Peptides

Improve activity, specificity, and more

Whether you have high specificity demands for API development, need to reduce product inhibition, or are concerned with unit economics for bulk production, Cradle allows you to set the most important optimization objectives and strike the balance between them.

Explore improvements beyond the active site

Beyond the active site, unintuitive positions in other protein domains can hold the key to large activity improvements. Cradle models are informed by the evolutionary context of your enzyme, as well as functional experimental data. You can choose to give Cradle freedom to introduce mutations anywhere in the protein sequence, or to define specific regions where you want to encourage or avoid mutations.

Improve performance of heterologous enzymes

Expressing an enzyme in a non-native host can compromise expression, and change the environmental context in which the enzyme needs to perform. Use Cradle to find designs that express well, and are functional given the temperature, pH, substrate availability in the host organism.

Screening data without the combinatorial headache

If you wish to start a project with variant libraries through site saturation mutagenesis or error-prone PCR, you can easily upload these datasets and use Cradle to identify the optimal recombinations of explored mutations, while identifying additional points in sequence space.

Leverage natural diversity

You can provide Cradle with diverse natural homologs of interest, helping the models tease out important underlying patterns.

Capture epistatic effects from the start of the program

From your first Cradle round, generate new proteins with multiple mutations in every variant. This ensures that you capture epistatic effects and explore the sequence space effectively.

Keep learning across assay conditions and experimental batches

As your enzyme evolves, you may increase the stringency of your assay in order to stay in its linear range. Similarly, experimental data can show batch effects when measurements are collected on different days or equipment. Cradle models learn throughout your project, connecting datasets even when the conditions under which they were created have been modified.

Enzymes

Binders

Peptides

Improve activity, specificity, and more

Whether you have high specificity demands for API development, need to reduce product inhibition, or are concerned with unit economics for bulk production, Cradle allows you to set the most important optimization objectives and strike the balance between them.

Explore improvements beyond the active site

Beyond the active site, unintuitive positions in other protein domains can hold the key to large activity improvements. Cradle models are informed by the evolutionary context of your enzyme, as well as functional experimental data. You can choose to give Cradle freedom to introduce mutations anywhere in the protein sequence, or to define specific regions where you want to encourage or avoid mutations.

Improve performance of heterologous enzymes

Expressing an enzyme in a non-native host can compromise expression, and change the environmental context in which the enzyme needs to perform. Use Cradle to find designs that express well, and are functional given the temperature, pH, substrate availability in the host organism.

Screening data without the combinatorial headache

If you wish to start a project with variant libraries through site saturation mutagenesis or error-prone PCR, you can easily upload these datasets and use Cradle to identify the optimal recombinations of explored mutations, while identifying additional points in sequence space.

Leverage natural diversity

You can provide Cradle with diverse natural homologs of interest, helping the models tease out important underlying patterns.

Capture epistatic effects from the start of the program

From your first Cradle round, generate new proteins with multiple mutations in every variant. This ensures that you capture epistatic effects and explore the sequence space effectively.

Keep learning across assay conditions and experimental batches

As your enzyme evolves, you may increase the stringency of your assay in order to stay in its linear range. Similarly, experimental data can show batch effects when measurements are collected on different days or equipment. Cradle models learn throughout your project, connecting datasets even when the conditions under which they were created have been modified.

Enzymes

Binders

Peptides

Improve activity, specificity, and more

Whether you have high specificity demands for API development, need to reduce product inhibition, or are concerned with unit economics for bulk production, Cradle allows you to set the most important optimization objectives and strike the balance between them.

Explore improvements beyond the active site

Beyond the active site, unintuitive positions in other protein domains can hold the key to large activity improvements. Cradle models are informed by the evolutionary context of your enzyme, as well as functional experimental data. You can choose to give Cradle freedom to introduce mutations anywhere in the protein sequence, or to define specific regions where you want to encourage or avoid mutations.

Improve performance of heterologous enzymes

Expressing an enzyme in a non-native host can compromise expression, and change the environmental context in which the enzyme needs to perform. Use Cradle to find designs that express well, and are functional given the temperature, pH, substrate availability in the host organism.

Screening data without the combinatorial headache

If you wish to start a project with variant libraries through site saturation mutagenesis or error-prone PCR, you can easily upload these datasets and use Cradle to identify the optimal recombinations of explored mutations, while identifying additional points in sequence space.

Leverage natural diversity

You can provide Cradle with diverse natural homologs of interest, helping the models tease out important underlying patterns.

Capture epistatic effects from the start of the program

From your first Cradle round, generate new proteins with multiple mutations in every variant. This ensures that you capture epistatic effects and explore the sequence space effectively.

Keep learning across assay conditions and experimental batches

As your enzyme evolves, you may increase the stringency of your assay in order to stay in its linear range. Similarly, experimental data can show batch effects when measurements are collected on different days or equipment. Cradle models learn throughout your project, connecting datasets even when the conditions under which they were created have been modified.

Got another use case? Let's talk.

Got another use case?
Let's talk.

If you can measure it, chances are Cradle's platform can help you progress faster. Talk to a Cradle expert to learn what Cradle can do for your organization.

Design in Cradle.
Validate in the lab.

Design with generative AI customized to your protein and run experiments like you always do.

Learn

Generate

Review

Import your data

To begin, upload wet lab data from screening or optimization experiments. If you're starting a project from scratch, simply define your goals and assays to get going.

Explore your data quality

Cradle automatically analyses your experimental data to determine data quality and AI’s ability to learn from it. This helps you to understand how to improve your workflows and assays to get to optimal sequences faster.

Models learn with you

Every time you upload lab results, Cradle gets better at understanding your protein. This means increasingly performant variants that get you closer to your goals— every round.

Learn

Generate

Review

Import your data

To begin, upload wet lab data from screening or optimization experiments. If you're starting a project from scratch, simply define your goals and assays to get going.

Explore your data quality

Cradle automatically analyses your experimental data to determine data quality and AI’s ability to learn from it. This helps you to understand how to improve your workflows and assays to get to optimal sequences faster.

Models learn with you

Every time you upload lab results, Cradle gets better at understanding your protein. This means increasingly performant variants that get you closer to your goals— every round.

Learn

Generate

Review

Import your data

To begin, upload wet lab data from screening or optimization experiments. If you're starting a project from scratch, simply define your goals and assays to get going.

Explore your data quality

Cradle automatically analyses your experimental data to determine data quality and AI’s ability to learn from it. This helps you to understand how to improve your workflows and assays to get to optimal sequences faster.

Models learn with you

Every time you upload lab results, Cradle gets better at understanding your protein. This means increasingly performant variants that get you closer to your goals— every round.

Always secure.
Scale anytime.

Always secure.
Scale anytime.

Privacy-first AI

Your experimental data is only used to train your organization's custom models.

SOC 2 compliant

Cradle is fully SOC 2 compliant with active monitoring and tight security protocols in place.

Single sign-on

Secure and simple account management with SSO support for Google and Microsoft.

Custom predictors

BETA

Use internally developed predictors right in Cradle and see outputs inside Reports.

Zero-maintenance

Fully managed AI infrastructure with optimized GPUs that scale with your needs.

Ready to start?
Talk to our team

report
report
round
round
data
data

© 2025 · Cradle is a registered trademark

Built with ❤️ in Amsterdam & Zurich

© 2025 · Cradle is a registered trademark

Built with ❤️ in Amsterdam & Zurich