Before creating an assistant, set up credentials for your LLM, STT, and TTS providers in Integration → Vault. The creation wizard will ask you to select a provider — credentials must exist first.
Create your first assistant
Navigate to Assistants
Select your LLM provider and model
Write your system prompt
{{variable}} syntax to inject dynamic values at runtime — caller name, account ID, or any context passed when a call is initiated.Add tools (optional)
A newly created assistant has no deployment attached. It will not handle live calls or web sessions until you configure at least one deployment under Configure assistant → Deployments.
Configure your assistant
After creation, open Configure assistant from the top-right of the assistant page. Configuration is organized into six areas:Prompt & Model
System prompt, model selection, temperature, token limits, and advanced LLM parameters.
Voice Pipeline
STT provider and model, VAD sensitivity, noise cancellation, EOS detection, TTS provider and voice, pronunciation rules.
Knowledge & Retrieval
Attach knowledge bases, set retrieval method (hybrid, semantic, text), top-K, score threshold, and reranking.
Tools
Knowledge retrieval, API request, endpoint invocation, call hold, and end-of-conversation tool types.
Deployments
Phone, web widget, web app, WhatsApp, and API deployment channels — each with its own voice and experience settings.
Webhooks & Analysis
Post-call webhooks to downstream systems and analysis pipelines for conversation scoring and custom metrics.
Prompt and model
The prompt and model configuration defines what your assistant knows, how it reasons, and how it generates responses.System prompt
The system prompt is your primary control surface. It sets the assistant’s persona, scope of knowledge, constraints, and tone. Well-written prompts are the single biggest lever for assistant quality. Dynamic variables — inject runtime context into your prompt using{{variable_name}} syntax. Variables are automatically detected from your prompt and can be populated via the SDK when initiating a call:
LLM model parameters
OpenAI / Azure OpenAI parameters
OpenAI / Azure OpenAI parameters
| Parameter | Default | Range | Notes |
|---|---|---|---|
| Model | gpt-4o | — | gpt-4, gpt-4o, gpt-4.1-mini, o3, o4-mini, and more |
| Temperature | 0.7 | 0–2 | Higher = more creative, lower = more deterministic |
| Top P | 1 | 0–1 | Nucleus sampling; use with temperature, not instead |
| Max completion tokens | 2048 | ≥1 | Caps response length |
| Frequency penalty | 0 | -2 to 2 | Reduces word repetition |
| Presence penalty | 0 | -2 to 2 | Encourages topic diversity |
| Stop sequences | — | up to 4 | Tokens that halt generation |
| Tool choice | auto | none / auto / required | Control when tools are called |
| Reasoning effort | — | low / medium / high | For o-series models only |
| Response format | — | text / json_object / json_schema | For structured output use cases |
Anthropic parameters
Anthropic parameters
| Parameter | Default | Range | Notes |
|---|---|---|---|
| Model | claude-opus-4 | — | Opus 4, Sonnet 4, Sonnet 3.7, Haiku 3.5 |
| Max tokens | 1028 | ≥1 | Caps response length |
| Temperature | 1.0 | 0–1 | — |
| Top P | — | 0–1 | Nucleus sampling |
| Top K | — | — | Top-K sampling |
| Stop sequences | — | — | Halt generation tokens |
| Extended thinking | — | JSON config | Claude’s built-in reasoning mode |
AgentKit — custom LLM backend
AgentKit — custom LLM backend
AgentKit replaces the built-in LLM with your own gRPC server. Rapida streams user speech transcripts to your server and synthesizes your text responses to audio in real time.
Your server receives a bidirectional
| Parameter | Notes |
|---|---|
| Server URL | host:port of your gRPC service (e.g. my-server.example.com:50051) |
| TLS certificate | Optional — path to CA cert for mutual TLS |
| Metadata | Key-value map passed as gRPC metadata headers |
Talk stream. Rapida handles all audio — VAD, STT, TTS, telephony. Your server only handles text in / text out.See the AgentKit guide for implementation examples with LangChain, CrewAI, and Anthropic Claude.Voice pipeline
The voice pipeline controls how audio is captured, cleaned, converted to text, and how text responses are spoken back. These settings directly affect latency, accuracy, and caller experience.Voice input — STT, VAD, and noise processing
STT providers: Deepgram, AssemblyAI, Azure Cognitive Speech, Google Speech, OpenAI Whisper, Cartesia, Sarvam AI| Audio setting | Default | Range | What it controls |
|---|---|---|---|
| STT language | multi | Provider-dependent | Primary transcription language; multi enables auto-detection |
| STT model | nova-3 | Provider-dependent | Accuracy vs. latency tradeoff per provider |
| VAD threshold | 0.6 | 0.5–1.0 | Speech detection sensitivity. Lower = more sensitive (may catch noise). Higher = stricter (may miss quiet speech) |
| EOS timeout | 700 ms | 500–4000 ms | Silence duration before the assistant treats a pause as end-of-turn. Increase for conversational pauses; decrease for snappy IVR-style responses |
| EOS backoff | 2 | 0–5 | Re-prompts (“Are you there?”) before ending session on silence |
| Noise cancellation | rn_noise | — | RNNoise background noise suppression applied before VAD and STT |
EOS timeout is the most impactful latency lever after model selection. At 700ms, the assistant waits 0.7 seconds of silence before responding. Reducing to 500ms cuts perceived latency noticeably but risks cutting off slow speakers mid-sentence.
Voice output — TTS and pronunciation
TTS providers: ElevenLabs, Cartesia, Deepgram Aura, OpenAI TTS, Azure Speech, Google Cloud TTS, PlayHT, Sarvam AI| Speaker setting | Default | What it controls |
|---|---|---|
| Voice ID | Provider-dependent | The specific voice model. Supports cloned brand voices on ElevenLabs and Resemble |
| Speed / emotion | Normal | Provider-specific; Cartesia exposes slowest → fastest and emotion controls (anger:high, positivity:low, etc.) |
| Sentence boundaries | .!?;:—… | Characters that flush a TTS chunk for immediate playback. Tuning these reduces time-to-first-audio |
| Conjunction break | 240 ms | Pause inserted at conjunctions (and, but, or). Adds natural rhythm to long sentences |
| Pronunciation dictionaries | — | Normalize how abbreviations, currencies, dates, URLs, and technical terms are spoken |
Knowledge and retrieval
Attach one or more knowledge bases to give your assistant access to documents, FAQs, product data, or any content indexed in Rapida.| Setting | Default | Range | Notes |
|---|---|---|---|
| Retrieval method | hybrid | hybrid / semantic / text | Hybrid combines vector similarity and full-text search — best default for most use cases |
| Top K | 5 | 1–20 | Number of document chunks retrieved per query. Higher = more context, higher latency |
| Score threshold | 0.5 | 0–1 | Minimum relevance score. Raise to reduce noise; lower to improve recall |
| Reranking | off | — | Cohere reranker re-scores retrieved chunks before passing to the LLM. Improves precision at the cost of a small latency increase |
Tools
Tools extend what your assistant can do mid-conversation without breaking the voice flow. The LLM decides when to call a tool based on its description — write clear, specific descriptions.Knowledge Retrieval
Query a Rapida knowledge base in real time. Returns the most relevant document chunks to the LLM as context.
API Request
Call any external HTTP endpoint mid-conversation — CRM lookups, inventory checks, booking APIs. Define the request schema and the LLM populates the parameters from conversation context.
Endpoint (LLM Call)
Invoke a Rapida endpoint — a separately configured LLM prompt — for specialised sub-tasks: classification, extraction, or complex reasoning offloaded from the main conversation model.
Put On Hold
Pause the call and play hold music while a backend process completes — useful when a lookup or action takes longer than a voice turn allows.
End of Conversation
Terminate the call programmatically when the assistant determines the conversation objective has been met.
Tool names must use only letters, numbers, and underscores (no spaces). The description is passed directly to the LLM — it determines when and whether the tool is called. Be specific: “Search the product knowledge base for pricing information” outperforms “Search knowledge base”.
Conversation experience
These settings control the runtime behaviour of a live session — what happens when the caller goes silent, how long sessions last, and what the assistant says at the start of a call.| Setting | Default | Range | Notes |
|---|---|---|---|
| Greeting message | — | — | Spoken immediately when a call connects. Supports {{variable}} for personalisation |
| Idle message | Are you there? | — | Spoken after the idle timeout expires without user input |
| Idle silence timeout | 30 s | 15–120 s (phone) | Time before the idle message triggers |
| Idle backoff | 2 | 0–5 | How many times the idle message repeats before ending the session |
| Max session duration | 300 s | 180–600 s (phone) | Hard cap on session length — protects against runaway calls |
| Error message | — | — | Spoken when the assistant encounters an unrecoverable error |
Webhooks and post-call analysis
Webhooks fire atconversation.begin, conversation.completed, and conversation.failed events. Deliver transcripts, metadata, and analysis results to your CRM, data warehouse, or alerting system in real time.
Analysis pipelines run after a conversation ends. They invoke a configured Rapida endpoint — typically an LLM prompt — against the conversation transcript to produce structured output: sentiment scores, intent labels, CSAT predictions, compliance flags, or any custom metric.
See Webhooks and Analysis for full configuration details.
Version control
Every change to your assistant’s prompt, model, or parameters creates a new version. Versions let you safely iterate without affecting live traffic.| Action | What it does |
|---|---|
| Create version | Saves a new draft with updated prompt/model config and a change description |
| Release version | Promotes the version to live — all active deployments switch immediately |
| Rollback | Re-release a previous version to revert a bad change |
Next steps
Set up a phone deployment
Connect a telephony provider and go live with inbound and outbound calling.
Add a knowledge base
Index documents, wikis, and data sources for retrieval during calls.
Configure webhooks
Stream call events and transcripts to external systems in real time.
Build a custom LLM backend
Use AgentKit to plug your own reasoning engine into Rapida’s audio pipeline.