List agents

Authorizations

Authorization

string

headerrequired

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Response

200 - application/json

string

Unique identifier for the agent

agent_name

string

Human-readable agent name

agent_type

string

Type of agent

agent_status

enum<string>

Current status of the agent

Available options:

seeding,

processed

created_at

string

Timestamp of agent creation

updated_at

string

Timestamp of last update for the agent

tasks

object[]

An array of tasks that the agent can perform

tasks.task_type

enum<string>

Type of task

Available options:

conversation,

extraction,

summarization,

webhook

tasks.tools_config

object

Configuration of multiple tools that form a task

tasks.tools_config.llm_agent

object

required

Configuration of LLM model for the agent task

tasks.tools_config.llm_agent.agent_type

enum<string>

default: simple_llm_agent

Available options:

simple_llm_agent,

knowledgebase_agent

tasks.tools_config.llm_agent.agent_flow_type

enum<string>

default: streaming

Available options:

streaming

tasks.tools_config.llm_agent.routes

object

Semantic routing layer

tasks.tools_config.llm_agent.routes.embedding_model

string

default: snowflake/snowflake-arctic-embed-m

Since we use fastembed all models supported by fastembed are supported by us

tasks.tools_config.llm_agent.routes.routes

object[]

These are predefined routes that can be used to answer FAQs, or set basic guardrails, or do a static function call.

tasks.tools_config.llm_agent.llm_config

object

LLM configuration

tasks.tools_config.llm_agent.llm_config.agent_flow_type

enum<string>

default: streaming

Available options:

streaming

tasks.tools_config.llm_agent.llm_config.provider

string

default: openai

tasks.tools_config.llm_agent.llm_config.family

string

default: openai

tasks.tools_config.llm_agent.llm_config.model

string

default: gpt-3.5-turbo

tasks.tools_config.llm_agent.llm_config.summarization_details

string | null

tasks.tools_config.llm_agent.llm_config.extraction_details

string | null

tasks.tools_config.llm_agent.llm_config.max_tokens

integer

default: 100

tasks.tools_config.llm_agent.llm_config.presence_penalty

number

default: 0

tasks.tools_config.llm_agent.llm_config.frequency_penalty

number

default: 0

tasks.tools_config.llm_agent.llm_config.base_url

string

default: https://api.openai.com/v1

tasks.tools_config.llm_agent.llm_config.top_p

number

default: 0.9

tasks.tools_config.llm_agent.llm_config.min_p

number

default: 0.1

tasks.tools_config.llm_agent.llm_config.top_k

number

default: 0

tasks.tools_config.llm_agent.llm_config.temperature

number

default: 0.1

tasks.tools_config.llm_agent.llm_config.request_json

boolean

default: false

tasks.tools_config.synthesizer

object

required

Configuration of Synthesizer model for the agent task

tasks.tools_config.synthesizer.provider

enum<string>

required

Available options:

polly,

elevenlabs,

deepgram,

styletts

tasks.tools_config.synthesizer.provider_config

object

required

tasks.tools_config.synthesizer.provider_config.voice

enum<string>

required

Name of voice

Available options:

Matthew

tasks.tools_config.synthesizer.provider_config.engine

enum<string>

required

Engine of voice

Available options:

generative

tasks.tools_config.synthesizer.provider_config.language

enum<string>

required

Language of voice

Available options:

en-US

tasks.tools_config.synthesizer.provider_config.sampling_rate

enum<string>

default: 8000

Sampling rate of voice

Available options:

8000,

16000

tasks.tools_config.synthesizer.stream

boolean

default: false

tasks.tools_config.synthesizer.buffer_size

integer

default: 150

tasks.tools_config.synthesizer.audio_format

enum<string>

default: wav

Available options:

wav

tasks.tools_config.transcriber

object

required

Configuration of Transcriber model for the agent task

tasks.tools_config.transcriber.provider

enum<string>

required

Identification provider for Deepgram

Available options:

deepgram

tasks.tools_config.transcriber.model

enum<string>

required

Available options:

nova-2,

nova-2-meeting,

nova-2-phonecall,

nova-2-finance,

nova-2-conversationalai,

nova-2-medical,

nova-2-drivethru,

nova-2-automotive

tasks.tools_config.transcriber.language

enum<string>

required

Available options:

en,

hi,

es,

fr

tasks.tools_config.transcriber.stream

boolean

default: false

tasks.tools_config.transcriber.sampling_rate

integer

default: 16000

tasks.tools_config.transcriber.encoding

enum<string>

default: linear16

Available options:

linear16

tasks.tools_config.transcriber.endpointing

integer

default: 100

tasks.tools_config.input

object

required

Configuration of Input handler

tasks.tools_config.input.provider

enum<string>

required

Available options:

twilio,

plivo

tasks.tools_config.input.format

enum<string>

default: wavrequired

Available options:

wav

tasks.tools_config.output

object

required

Configuration of Output handler

tasks.tools_config.output.provider

enum<string>

required

Available options:

twilio,

plivo

tasks.tools_config.output.format

enum<string>

default: wavrequired

Available options:

wav

tasks.tools_config.api_tools

object

Api tools you'd like the agents to have access to

tasks.tools_config.api_tools.tools

object[]

Description of all the tools you'd like to add to the agent. It needs to be a JSON string as this will be passed to LLM.

tasks.tools_config.api_tools.tools.name

string

Any unique name for this function tool

tasks.tools_config.api_tools.tools.key

enum<string>

default: transfer_call

Available options:

transfer_call

tasks.tools_config.api_tools.tools.description

string

Use this tool to transfer the call

tasks.tools_config.api_tools.tools.parameters

object

tasks.tools_config.api_tools.tools_params

object

Parameters for each tool, where keys must match the name field in the tools array.

tasks.tools_config.api_tools.tools_params.transfer_call_support

object

tasks.toolchain

object

Agent will execute these tools in the specified order

tasks.toolchain.execution

enum<string>

required

Available options:

parallel,

sequential

tasks.toolchain.pipelines

enum<string>[]

required

Available options:

transcriber,

llm,

synthesizer

tasks.task_config

object

Should be used onkly in conversation task for now and it consists of all the required configuration for conversational nuances

tasks.task_config.hangup_after_silence

integer

default: 10

Time to wait in seconds before hanging up in case user doesn't speak a thing

tasks.task_config.incremental_delay

integer

default: 400

Since we work with interim results, this will dictate the linear delay to add before speaking everytime we get a partial transcript from ASR

tasks.task_config.number_of_words_for_interruption

integer

default: 2

To avoid accidental interruption, how many words should we wait for before interrupting

tasks.task_config.hangup_after_LLMCall

boolean

default: false

Weather to use LLM prompt to hang up or not. Pretty soon this will be replaced by predefined function

tasks.task_config.call_cancellation_prompt

string

tasks.task_config.backchanneling

boolean

default: false

This will enable agent to acknowledge when user is speaking long sentences

tasks.task_config.backchanneling_message_gap

integer

default: 5

Gap between every successive acknowledgement. We will also add a random jitter to this value to make it more random

tasks.task_config.backchanneling_start_delay

integer

default: 5

Basic delay after which we should start with backchanneling

tasks.task_config.ambient_noise

boolean

default: false

Toggle to add ambient noise to the call to add more naturalism

tasks.task_config.ambient_noise_track

enum<string>

default: office-ambience

Track for ambient noise can be coffee-shop, call-center, office-ambience

Available options:

office-ambience,

coffee-shop,

call-center

tasks.task_config.call_terminate

integer

default: 90

The call automatically disconnects reaching this limit

agent_prompts

object

Prompts to be provided to the agent. It can have multiple tasks of the form task_<task_id>

agent_prompts.task_1

object

agent_prompts.task_1.system_prompt

string

required

The system prompt fed into the agent

API Documentation

Current APIs

Legacy APIs

Authorizations

Response