Clinical and Bioinformatics Solutions for Genomics Labs
From order intake to variant interpretation, purpose-built solutions covering the full lab workflow, compliant with 21 CFR Part 11, HIPAA, and CLIA/CAP standards.
Where we fit in your workflow
We've built a suite of genomics-native products that labs deploy to eliminate manual bottlenecks, reduce operational overhead, and accelerate time-to-report.
Solutions built for genomics labs and diagnostic organizations
Missed orders, late reports, invisible pipeline failures, interpretation backlogs, these aren't edge cases, they're structural. We built these solutions because labs doing critical work deserve infrastructure that keeps up with them.
SmartReq
Intake & Operations
Intelligent patient requisition form extraction
Challenge
Intake teams manually re-key hundreds of requisition forms every day across dozens of referring provider layouts. A wrong insurance ID, a missing ICD code, a transposed patient DOB, errors introduced at intake propagate all the way to billing and cause claim denials. There is no scalable manual fix.
Key Features
- 20+ referring provider form templates supported, with template detection at intake
- Reads printed text, handwritten entries, checked boxes, and circled options
- Schema validation before human review, distinguishes a clean extraction from a valid one
- Low-confidence fields flagged explicitly, never silently dropped
- On-premise deployment, patient data never leaves your infrastructure
Intergenix
EHR Integration
EHR order validation and management layer that sits over Mirth Connect
Challenge
A meaningful share of EHR orders coming through Mirth Connect fail silently, duplicate submissions from retries, missing insurance segments, malformed HL7. These failures are discovered reactively: billing denials, downstream rejections, or a clinic calling to ask why a result hasn't arrived. Operations teams have no real-time visibility and no structured path to resolution.
Key Features
- Drools-based rules engine validates every order at intake, completeness, validity, and business rules
- Deduplication runs before validation, duplicate orders never enter the validation queue
- Field-level actionable status per order, specifies exactly which fields failed and why
- Structured clinic outreach workflow with follow-up tracking and escalation path
- Pattern detection, surfaces which clinic or EHR vendor is the root cause of recurring failures
StrixFlow
Pipeline Monitoring & Orchestration
Real-time bioinformatics pipeline observability, compute cost tracking, and compute management
Challenge
Pipeline failures are discovered at the clinical end, not the pipeline end. Labs running Nextflow or Snakemake on HPC or cloud have no unified view of what is running, what failed, and why. Compute costs are invisible until the cloud bill arrives. Existing tools like Terra and Seqera Tower solve parts of this, but not cost-per-sample visibility, step-level failure diagnosis, and HPC/cloud unified monitoring together in a self-hosted deployment the lab controls.
Key Features
- Multi-workflow support: Nextflow DSL2, Snakemake, WDL, CWL, connects without pipeline rewrites
- Step-level DAG visualization with failure diagnosis and monospace error output per failing process
- Cost-per-sample tracking across HPC (SLURM, PBS, LSF) and cloud environments in one unified view
- Pipeline Reproducibility Manifests, tool versions, container digests, and parameter sets captured per run
- RBAC across roles with 21 CFR Part 11 compliant immutable audit logging
Varion
Clinical Genomics
End-to-end variant interpretation with ACMG/AMP classification and 21 CFR Part 11 audit trails
Challenge
Analysts manually query ClinVar, gnomAD, and OMIM for every candidate variant and apply ACMG criteria by hand. At scale, this becomes the primary TAT bottleneck. Annotation API costs compound fast with no local caching strategy, and there is no structured reasoning layer to build AI-assisted interpretation on top of.
Key Features
- Three parallel scoring tracks: ACMG/AMP rule engine, HPO phenotype similarity, AI pathogenicity scoring
- Four-tier local annotation cache (Redis, Postgres/PGVector, DuckDB, live API fallback), no per-query cost at scale
- Germline and somatic schemas sharing a GA4GH VRS-compliant canonical variant store
- Human-in-the-loop sign-off, AI assembles evidence, analyst decides, no report issues without approval
- FHIR/HL7 integration for downstream EHR and LIMS reporting
GENVAR
Knowledge Management
Proprietary variant KMS with GA4GH VRS canonical store and PGVector AI co-pilot layer
Challenge
A lab's most valuable asset is its accumulated clinical judgment, but when that knowledge lives in a vendor platform like Emedgene, the lab doesn't own it. Classification reasoning, reclassification history, and evidence trails sit in a form that can't be queried, migrated cleanly, or used as a foundation for AI-assisted interpretation. When analysts leave, their reasoning walks out with them. When ClinVar updates, nobody finds out until a clinician asks.
Key Features
- GA4GH VRS-compliant canonical variant identity, same variant across germline and somatic workflows linked to one record
- Every classification is a versioned event, reclassification creates a new version, prior records never overwritten
- PGVector semantic search across the full institutional interpretation history
- AI Interpretation Co-pilot retrieves most relevant internal precedents during analyst review via RAG
- ClinVar diff monitoring, reclassification candidates surfaced proactively on every new release
ReportStudio
Lab Operations
Drag-and-drop clinical report builder with template versioning and approval workflows
Challenge
Every new report type, panel update, or regulatory language change requires a developer ticket and a 3–4 month cycle. Lab staff have no direct control over the tools they depend on daily. Without a formal approval layer, unapproved templates can enter production and unsigned reports can reach clinicians, creating accreditation exposure the lab may not even be aware of.
Key Features
- Drag-and-drop canvas for composing report layouts, no code required
- Reusable component library: headers, result tables, interpretation blocks, dynamic variable placeholders
- Full template version history with approval workflow, no template enters production without sign-off
- Pipeline JSON output maps directly to report placeholders, no copy-paste, no reformatting step
- 21 CFR Part 11 compliant audit trail per template and per issued report
Powering Genomic Labs, Delivering Measurable Impact
Genomics labs lose thousands of hours a year to manual workflows that were never designed to scale. These numbers show what gets recovered.
We fit into your world, not the other way around
Your stack, your infrastructure, your compliance requirements. We've never asked a lab to rebuild what's already working. We come in where the gaps are and leave with something your team owns.
See which solution fits your workflow
15-minute scoping call. We'll map your current stack and tell you exactly where we can help, or where we can't.