Brilo AI does not currently offer a global switch from workflow-style integrations to free-form natural-language prompts. Brilo AI voice agents rely on configured workflows, routing rules, and knowledge sources to ensure predictable, auditable behavior. Support for direct natural-language prompting (free-form prompts) is an active product question and may be enabled selectively when integrated with controlled knowledge and intent pipelines. If you need conversational prompting sooner, Brilo AI can often approximate natural-language flows by extending existing workflows, adding richer NLU models, and expanding your knowledge base. Contact your Brilo AI account team to discuss roadmap timing, pilot options, or a tailored implementation plan.

When will Brilo AI accept natural-language prompts? — Brilo AI currently uses workflow-style integrations for production voice agents; natural-language prompting is not generally available and requires planning with your account team.

Will Brilo AI let agents use free-form prompts instead of workflows? — Brilo AI can be configured to accept controlled free-form inputs in pilot or custom solutions; reach out to your implementation manager for options.

Can I switch my Brilo AI voice agent from workflows to prompts today? — Not as a one-click global change; Brilo AI recommends iterative rollout using staged NLU and knowledge updates.

Buyers ask because natural-language prompts promise faster iteration and more flexible agent behavior than predefined workflow-style integrations. Enterprise teams worry about answer consistency, auditability, and regulatory risk when moving from deterministic workflows to free-form prompt inputs. Brilo AI customers need clarity on how a move to natural-language prompting would change routing, logging, and compliance controls before committing to it in production.

Brilo AI voice agents currently run using workflow-style integrations: configured flows, explicit routing, and connected knowledge sources drive every call. When you ask Brilo AI to extend beyond fixed flows, Brilo AI evaluates three areas: natural-language understanding (NLU) accuracy, knowledge retrieval (semantic search), and safe escalation points to humans.

In Brilo AI, workflow-style integrations are a configured sequence of prompts, business-rule routing, and API/webhook calls that define how the voice agent handles a call.

In Brilo AI, natural-language prompts are an input mode where callers or operators provide free-form text or speech that the Brilo AI NLU interprets to choose actions or replies.

In Brilo AI, knowledge base is the structured and unstructured content set (documents, FAQs, CRM fields) Brilo AI consults for answers and retrieval.

Technical terms in scope: NLU, natural-language understanding, prompt engineering, intent detection, semantic search, webhook, CRM.

Brilo AI applies guardrails to keep behavior predictable and auditable when prompting is considered. Typical guardrails include strict confidence thresholds for NLU-based actions, required fallback to workflow nodes when confidence is low, logging of prompt inputs, and enforced routing to human agents for high-risk intents.

In Brilo AI, confidence threshold is a configurable setting that determines when an NLU-derived action is allowed to execute versus when the agent falls back to a workflow or escalation.

Brilo AI will not remove audit trails, change routing silently, or act on prompt-derived actions when regulatory or high-risk flags are present unless explicitly configured to do so. Data governance controls limit how prompt text is stored and who can access training examples.

Healthcare example: A hospital uses Brilo AI voice agents to triage appointment requests. Under workflow-style integrations, the agent follows a scripted flow to collect patient identifiers and scheduling preferences. If natural-language prompts are enabled in a controlled pilot, Brilo AI can accept a free-form patient sentence like “I need a cardiology follow-up next week” and map it to intent (specialty, urgency, date range) while still requiring confirmation steps and human review for clinical escalation.

Banking / Financial Services example: A bank uses Brilo AI for account balance and dispute routing. With workflow-style integrations, specific API calls return account status. In a constrained prompt mode, callers can say “I noticed an unauthorized charge” and Brilo AI’s intent detection maps to dispute workflows, triggers required authentication checks, and creates a ticket in your CRM while logging every step for compliance review.

Insurance example: An insurer configures Brilo AI to collect claim starter information. Prompts can let claimants describe damage in their own words; Brilo AI extracts structured fields (date, policy number, loss type) and routes to adjuster workflows while preserving the original transcript for audit.

Brilo AI supports deterministic handoff paths that you configure. When NLU confidence is below the configured threshold or when a policy flag triggers, Brilo AI voice agent can:

Handoff behavior is configurable per flow: you decide whether to pass the full transcript, extracted structured fields, or only a summary. Brilo AI preserves logs and timestamps so human agents see why a handoff occurred and what the agent already collected.

When Brilo AI is extended to controlled natural-language prompting, realistic outcomes include faster conversational flows for common requests, fewer menu-driven dead-ends for callers, reduced manual form entry for agents, and higher capture of structured data from free-form speech. The main operational gains come from better first-contact resolution and fewer unnecessary transfers when confidence and guardrails are tuned appropriately.

Does Brilo AI currently support free-form natural-language prompts in production?

Brilo AI primary production deployments use workflow-style integrations for predictable, auditable behavior. Free-form prompt capabilities are handled via pilots or custom engagements and require configuration of NLU, confidence thresholds, and governance controls.

Will natural-language prompting change my audit and logging practices?

Yes. Enabling prompting increases the need to log raw transcripts, NLU confidence scores, and mapping decisions. Brilo AI preserves these artifacts and supports configurable retention and access controls to align with your governance policies.

How does Brilo AI prevent incorrect actions from a misinterpreted prompt?

Brilo AI uses confidence thresholds and fallback workflows. If the NLU confidence is low or the intent is high-risk, the agent follows a preconfigured escalation path to human agents or confirmation steps instead of executing the action automatically.

Can Brilo AI extract structured data from a free-form prompt?

Yes when configured: Brilo AI can extract entities and map them to structured fields for CRM records or tickets, provided you supply representative training data and configure the extraction rules.

What integrations are required to enable prompting safely?

Typical integrations include your CRM for record creation, a webhook endpoint for custom workflows, and your knowledge sources for semantic retrieval. Brilo AI works with these endpoints to validate and act on extracted entities while preserving audit logs.