Screen-aware prompts
Ask AI receives compact context from the active page: page type, visual organization, current task, and scientific domain. That keeps answers grounded in the workflow the researcher is already using.
Ask AI
Private AI help from the lab floor
Aspis Bio brings screen-aware AI to Android and iPhone so researchers can ask quick scientific questions, check calculations, and summarize lab context through server-side routing, provider ZDR controls, and no provider keys inside the app.
Server-side routing
Aspis Bio Workers select models, enforce quotas, stream responses, and route through Cloudflare AI Gateway. Upstream providers receive model requests without the user's app account identity attached.
ZDR-aware providers
Nebius and Groq support zero-data-retention controls for inference. Aspis Bio is designed to use those controls where available, while keeping Cloudflare as the visible routing and usage layer.
PII reduction layers
Aspis Bio uses deterministic stripping rules, not ML Kit, for current PII reduction: local export sanitization removes device file paths, Worker-side search sanitization removes obvious email, phone, and sample ID patterns, and Cloudflare AI Gateway DLP can scan AI requests and responses when enabled.
Research boundaries
Ask AI is for research support: bench math, qPCR checks, biology context, chemistry calculations, and workflow guidance. It is not for clinical decisions or regulated human-genomics decision-making.
Flow
From question to streamed answer
1User asks from the current Android or iPhone screen.
2The app and Worker reduce obvious local identifiers where the flow supports it.
3The Worker selects the model, applies limits, and routes through Cloudflare AI Gateway.
4Gateway DLP can flag or block sensitive request or response content when enabled.
5Nebius or Groq answer without receiving the user's app account identity.