Computational Biology Consultancy
Turning the AI revolution into your competitive advantage.

Bespoke AI/ML-driven protein design, tooling services, technical advisory, and next-generation in silico experimentation
AI tools don't fail because they're broken.
They fail because no one is responsible for making them learn. Fine tuning. Methodical configuration.
In protein design, teams using the same tools routinely get dramatically different results. The difference isn't stochasticity — it's whether someone owns the learning loop. The group that succeeds knows how to interpret model outputs in the context of biochemical constraints, make iterative design decisions when initial results do not converge, and navigate the inherent uncertainty of computational protein design where even ideal-profile targets can fail. And then feed experimental results back into the next cycle.
The most credible successes in ML-driven biology aren't platforms or open-source models. They are specialized learning systems operated by cross-functional experts that are deeply integrated with real experiments — both in the lab and in silico — and get smarter with every cycle. Leviathan Labs owns both the operational runtime and the wet lab data integration that make that possible.
This is why services will outperform tools in this domain for the foreseeable future. Open-source tools and AI platforms are powerful and necessary — but rarely sufficient for protein design without the cross-domain expertise to customize them to real workflows and own the learning over time. Leviathan Labs was built around this reality.
Your boutique AI department for protein engineering.
We embed into your pipeline as the cross-domain AI department that tools can't provide. We own the learning loop driving better outcomes from generative protein modeling and drive convergence between wet-lab results and computational modeling. Every campaign we run is customized to your specific target, your assay constraints, and the specifications of your customers.
Our expertise spans the full computational stack: generative structure-based design, physics-based validation, and iterative refinement informed by wet-lab results. When a design cycle doesn't converge, we don't wait — we diagnose, adjust, and re-run. That capacity to operate under uncertainty is what separates a productive ML-driven biology program from one that stalls.
The result is a system that gets smarter with every cycle and a client that retains the results it generates. We characterize your target, tune the design approach, and deliver clear data science reports for every campaign: what was tested and the results justifying delivery.
End-to-end computational campaigns
tailored to your target of interest.
Protein Design
De novo protein design and sequence optimization using generative AI models, multi-round lead optimization, and energy-based scoring.
Antibody Engineering
Computational antibody discovery, affinity maturation, humanization, and developability assessment for therapeutic candidates.
Campaign Operations
End-to-end computational protein design campaigns — from target briefing to candidate delivery. No dashboards to learn, no pipelines to maintain. Just results.
Bespoke Tooling
Off-the-shelf models rarely fit real workflows out of the box. We fine-tune, modify, and experiment to deliver custom computational solutions built around your specific targets, data, and pipeline requirements.
Model Fine Tuning
Fine tune protein language models, inverse folding models and generative models on your proprietary data to improve hit rates.
Technical Advisory
Building a computational biology capability is a different problem than buying one. We advise biotech leadership on how to evaluate, structure, and scale AI-driven protein design programs.
Not a generalist shop.
A technical partner.
Deep Technical Specialization
We focus exclusively on computational protein engineering. No generalist outsourcing — every engagement is led by domain experts.
End-to-End Campaigns
From target analysis through delivery of AI-optimized sequences — tailored to your specific pipeline and goals.
Boutique Model, Senior Access
Direct access to senior expertise on every project. No layers of project managers — faster iteration, clearer communication.
Partnership, Not Vendor
We embed with your team as collaborators, not transactional service providers. Your success is our metric.
Built by scientists and engineers.
Powered by deep domain expertise.

Grant Armstrong
Founder and Computational Biologist
Click to view bio
Grant Armstrong
Specialized in computational protein design and molecular modeling, Grant brings 6 years of expertise in leveraging AI-driven approaches to solve complex structural biology challenges. His work spans de novo protein engineering, antibody optimization, and physics-based validation pipelines, enabling your solution to leverage both data driven and biophysical approaches.
LinkedIn →
Andrew Gallagher
Full Stack Software Engineer
Click to view bio
Andrew Gallagher
Andrew architects and builds the computational infrastructure that powers Leviathan Labs’ protein design workflows. From high-throughput data pipelines to interactive analysis tools, he ensures that cutting-edge science is backed by robust, scalable software systems.
LinkedIn →Ready to accelerate your
protein engineering pipeline?
We work with companies pushing the boundaries of protein therapeutics. Let's discuss how computational design can advance your program.
Or email us at grant.armstrong@leviathanlabs.co — LinkedIn