Every layer feeds the next.
Results move from sample to signal to analysis without being exported, reformatted, or re-entered along the way.
The Convah Platform
From sample to answer, built as one stack.
Convah is a vertically integrated platform for pathogen intelligence: sample handling, molecular detection, and AI analysis designed as one system, not a set of tools you stitch together.
Why integration
A suite of disconnected tools isn’t a platform. An integrated stack is.
The data compounds.
Every layer shares one data model, so each run is designed to make the next one faster and better informed.
One owner of the answer.
One system means one audit trail, from the moment a sample is collected to the result you report.
The Convah stack
Four layers. One system.
The analysis layer, available now
From raw reads to answers, in one conversation.
The Convah Platform runs nine analytical engines in one shared environment, behind a single natural-language interface. Describe the question, and it builds the pipeline, runs it, and shows its work.
01 / 09
Upload
Paired-end FASTQ or raw data from Illumina, Oxford Nanopore, and PacBio, with automatic quality assessment.
IlluminaOxford NanoporePacBio
02 / 09
Assemble
De novo genome assembly with automatic contig quality metrics, N50 scores, and completeness assessment.
MEGAHITmetaSPAdes
03 / 09
Classify
Species-level taxonomic classification across more than 1.8 million reference genomes, typically in under 15 minutes per metagenomic sample.
Kraken2
04 / 09
Detect
More than 200 known pathogens, including MRSA, VRE, and CRE, flagged with clinical severity scores and abundance.
Pathogen panelSeverity scoring
05 / 09
Extract
Target organism reads pulled from complex metagenomic mixtures by taxonomy, abundance, or severity.
Taxon filters
06 / 09
Phylogeny
Interactive evolutionary trees in rectangular, radial, and linear views, with branch lengths and bootstrap values.
Tree builder
07 / 09
Dock
AI structure prediction combined with molecular docking for protein-ligand binding analysis.
BoltzLightDock
08 / 09
Evolve
Codon-based mutation libraries with automated binding affinity scoring across all variants.
Evolve Engine
09 / 09
Score
Combined interaction scoring ranks complexes and variants for review.
Score aggregation
No command line.
Every tool runs through a natural language interface. No scripting, no configuration files, no terminal.
No scripting.
Complex multi-step pipelines assembled automatically from your plain-English description of the goal.
No coding required.
Kraken2, MEGAHIT, metaSPAdes, LightDock, Boltz: all accessible to any researcher, any background.
Plan your project
Plan the work, and the budget, before you run it.
Whether you are writing a grant or planning a commercial program, start with a few directional inputs or upload your own data. The estimator projects jobs, compute, cost, and turnaround for a grant period or a monthly operation.
Try the calculator
Plan a grant budget or a commercial program with the public calculator.
Open the calculatorIllustrative preview. Figures shown are not estimates or quotes.
Applications
One stack, many applications.
- Software and compute together, planned against grant budgets
- No coding required for students and postdocs
- One shared environment for the whole lab
- Protein analysis, docking, and the Evolve Engine in one place
- Plain-English interface with no command line
- Built-in visualization and export
- One request replaces a multi-tool workflow
- Concurrent jobs with live status
- Standardized, reproducible client reports
- MRSA, VRE, and CRE detection from metagenomic samples
- Severity scoring and abundance quantification
- Results in hours rather than days
For research use only. Not for use in diagnostic procedures.
- Threat identification with the Convah Evolution module
- Severity-scored alerts and phylogenetic tracing across hosts and sites
- Reports structured for agency and partner review
Our origin
It started with deer and elk.
Convah’s work began with an early breakthrough in studying the health of cervid populations: deer and elk.
Getting there meant rethinking every step at once: how a sample is handled, how an early signal is read, and how the results are analyzed. Fixing one step was never enough. That lesson, and the breakthroughs that followed, became the integrated stack.
3 in 4
new or emerging infectious diseases in people come from animals, which is why animal health remains a proving ground for the platform.
Evidence first.
Peer-reviewed methods
Our engines run established, published tools rather than black boxes, so results can be checked against the literature.
Reproducible by default
Every run records its tool versions, parameters, and reference data. Rerun it next year and get the same answer.
Validated before claimed
We publish performance data with our partners before we make claims about it.
Scientists and builders.
Molecular biology, infectious disease, genomics, and AI engineering, building one stack.
Company
Key advisors
Get in touch.
Sign up todayGet access to the Convah Platform.