API Documentation
Built-in Tools
Custom Tools
Calls
Post-call Analysis
Phone Numbers
- GETRetrieve a paginated list of all phone numbers configured in your PIA Platform account
- POSTCreate a new phone number
- GETRetrieve detailed information about a specific phone number configuration
- DELRemove a phone number configuration from the PIA Platform
- PATCHUpdate the configuration of an existing phone number
Knowledge Bases
Public Sharing
List AI Agents
List AI agents with pagination.
Response
List of AI agents matching the query criteria
Name of the AI agent
Unique identifier for the AI agent
ID of the user who created this AI agent
Language model configuration for the AI agent
Unique identifier for the Language Learning Model (LLM)
Indicates this is a simple LLM type that uses standard model providers
simple Name of the specific model from the provider (e.g. 'gpt-4', 'claude-2')
The provider of the language model - OpenAI, Anthropic, or Llama3
openai, anthropic, llama3 The base prompt that defines the LLM's behavior, role and operating parameters
Controls randomness in model outputs - 0 is focused/deterministic, 1 is more creative/random
0 < x < 1List of dynamic data fields that must be provided to this LLM during operation
List of tools/functions available to the LLM for external interactions and API calls
Speech-to-text configuration for the AI agent
Specific Deepgram model being used for speech recognition, each optimized for different use cases
nova-2-general, nova-2-phonecall, nova-meeting, nova-2-meeting, nova-2-finance, nova-2-conversationalai, nova-2-voicemail, nova-2-video, nova-2-medical, nova-2-drivethru, nova-2-automotive, enhanced-general, enhanced-meeting, enhanced-phonecall, enhanced-finance, base, meeting, phonecall, finance, conversationalai, voicemail, video, whisper-tiny, whisper-base, whisper-small, whisper-medium, whisper-large Indicates Deepgram is being used as the STT provider
deepgram Text-to-speech configuration for the AI agent
Indicates ElevenLabs is being used as the TTS provider
elevenlabs Unique identifier for the selected ElevenLabs voice
Human-readable name of the selected voice
Specific ElevenLabs model version being used for voice synthesis
eleven_turbo_v2_5, eleven_turbo_v2, eleven_multilingual_v2, eleven_monolingual_v1 Controls variation in voice output - higher values increase variability
0 < x < 1Type of ambient background sound to use during calls
office, cafe, restaurant, park, street, home, library, airport, train_station, beach, none Volume level for ambient sound, from 0 (muted) to 2 (loud)
0 < x < 2Configuration for dynamic data sources that can be used during calls
Request body data to send to the endpoint
Whether to cache the response data for subsequent use
HTTP headers to include in the request
HTTP method to use when calling the endpoint (GET, POST, etc)
URL query parameters to include in the request
Configuration for parsing and using the endpoint response data
Additional context or metadata about this data field
The actual data value retrieved from the external source
Identifier for this dynamic data field
Maximum time in seconds to wait for the endpoint response
Endpoint URL for retrieving dynamic data
Whether to allow users to interrupt the AI agent while speaking
Number of seconds of silence before automatically ending the call
10 < x < 30ID of the associated knowledge base, if any
Language code for the AI agent in BCP-47 format
Maximum duration of a call in seconds
30 < x < 3600Minimum duration of speech in seconds before interruptions are allowed
0 < x < 15Minimum number of words that must be spoken before interruption is allowed
0 < x < 5Configuration for analysis items to be processed after calls
Detailed description of what this analysis item measures or determines
Identifier for this analysis item
Prompt used to guide the AI in performing this analysis
Indicates this analysis item returns a text string
string Example responses to guide the analysis
Time in seconds to wait for voicemail detection
x > 0Time in seconds to wait before considering speech has ended
0 < x < 15URL to send webhook notifications about call events
Custom welcome message to play at the start of calls
Token to retrieve the next page of results. Null if there are no more results