— Platform · AI →
An assistant that reads your real data without it ever leaving the platform.
Not a chatbot with a logo. Holonyx AI knows the module structure and data model, offers actions that map to real platform operations, and works under the same permissions as the person asking.
Holonyx AI
A daily brief and actions that map to real operations.
Your tasks, active work and what needs attention in one card, plus suggested actions that scaffold or run real platform operations instead of generic prompts.
Suggested actions
Analyze Experiment
Flag deviations against acceptance criteria on request.
Draft Batch Record
Scaffold a batch record from an approved recipe.
Run DOE Design
Turn ingredient ranges into a designed run matrix.
Compliance Check
Check a record or procedure against Part 11 expectations.
Build Workflow
Generate a BPMN workflow from a description.
Search Literature
Look up published work relevant to a question.
Your model, your choice
A platform-hosted model or a commercial one — the provider is an admin setting.
Ollama or LM Studio run as a model hosted inside the platform — isolated to your organization, never sent to a third-party AI vendor — or configure Anthropic, OpenAI, Google, Mistral or Cohere instead. Set a default and a fallback, test each connection, and watch usage per provider.
- A platform-hosted model keeps prompts isolated to your organization, never sent to an outside AI vendor
- A prompt library of 500+ curated pharmaceutical and scientific prompts
- The system prompt is a configurable parameter under admin control
— How to use it responsibly →
AI output is advisory. It is reviewed by qualified staff before it is used in any GxP decision, and it works through the same role permissions and API paths as a person — it cannot see or do what the user cannot.
Ask it about your own data.
A working session with the team that built it — your modality, your compliance regime, your instruments.