Skip to main content

OUR PLATFORM

Three layers. One scientific platform.

Hordago separates product shells, scientific engines, and execution infrastructure. This allows new domain products to be generated without rebuilding the stack from scratch.

Platform Architecture

Built in Layers

Domain Operating Systems

CRISPRos CRISPR Operating System
GWASos GWAS Operating System

Shared Scientific Engines

biocontext7 Tool discovery
paper-intelligence Literature engine
co-scientist Hypothesis engine
co-writer Evidence-locked writing
Knowledge Graph Entity + provenance
skill-discovery-mcp Tool discovery

Factory & Execution

life-sciences-factory Scientific compiler
specforge Spec generation
velocity Batch execution

LAYER 2

Shared Scientific Engines

The reusable substrate that powers every DomainOS product.

co-writer

Scientific Writing Engine

Evidence-locked manuscript drafting. 34 skills, 5 agents. Artifacts, not hallucination.

co-scientist

Hypothesis Discovery Engine

6-phase pipeline with Elo tournament ranking for hypothesis selection. 8 skills, 7 agents.

sophia-claudette

Epistemic Reasoning Engine

Uncertainty calibration, causal skepticism, drift auditing. 7 skills, 9 agents.

paper-intelligence

Claim Extraction Engine

Structured claim graphs, contradiction detection, citation networks. 12 skills, 3 agents, 5 MCP.

knowledge-graph

Biomedical Knowledge Graph

Neo4j + DuckDB + FTS5 across 7 namespaces. 5 skills, 1 agent, 11 MCP tools.

biocontext7

Deep Skill Bundles

2,000+ hash-verified bioinformatics skill bundles served via MCP. External MCP server.

LAYER 3

Factory & Execution

How Hordago manufactures scientific products.

Scientific Compiler

Deterministic Build System

Spec Engine

Work Decomposition

Agentic Orchestration

Parallel Execution

THE FLYWHEEL

How DomainOS products learn

Every experiment makes the system smarter.

1

Prioritize

Foundation models score rare variants and predict regulatory effects with calibrated uncertainty

Variant scoring
2

Plan

Agentic workflows generate optimized CRISPR perturbation strategies and protocols

Guide design
3

Execute

Experiments run with full provenance capture, tracking every parameter and condition

Lab execution
4

Learn

Outcomes feed back to improve predictions, closing the loop for continuous improvement

Model updates
Continuous Learning Cycle

SECURITY & COMPLIANCE

Built for regulated environments.

Enterprise-grade security and compliance infrastructure designed for life sciences research.

Data Privacy

All data processing with your own API keys — nothing stored on our servers. Your research data never leaves your infrastructure.

Full Provenance

Every output includes a cryptographic manifest: inputs, git commits, tool versions, environment, and SHA-256 hashes for complete auditability.

Access Controls

Role-based access with human approval gates at every critical decision point. AI proposes, scientists decide.

SOC 2 Type II IN PROGRESS
HIPAA-Ready ARCHITECTURE

Ready to accelerate your research?

Learn how our platform can transform your scientific workflows.