SDK Overview

Official SDKs for the Burki Voice AI Platform - Python, JavaScript/TypeScript, and Go

Burki provides official SDKs in Python, JavaScript/TypeScript, and Go to help you integrate voice AI capabilities into your applications quickly and easily.

Available SDKs

Language	Package	Installation	Status
Python	burki	pip install burki	Available
JavaScript/TypeScript	@burki.dev/sdk	npm install @burki.dev/sdk	Available
Go	github.com/burki-ai/burki-go	go get github.com/burki-ai/burki-go	Available

---

Feature Support Matrix

All SDKs provide comprehensive access to the Burki API:

Feature	Python	JavaScript	Go
Assistants	Full CRUD	Full CRUD + extras	Full CRUD
Calls	List, transcripts, recordings, metrics	Full management + analytics	Full management
Phone Numbers	Search, purchase, assign	Full management + webhooks	Search, purchase, assign
Documents (RAG)	Upload, list, status	Upload (file/URL), reprocess	Upload (file/URL)
Tools	Create, assign, Lambda discovery	Full CRUD + Lambda discovery	Create, assign, Lambda
SMS	Send, conversations	Full management + queue	Send, conversations
Campaigns	Create, start, progress	Full CRUD + contacts	Create, start, progress
Real-time Streaming	Async context managers	Async iterators	Channel-based
Async Support	Sync + Async methods	Native async/await	Goroutines
TypeScript Types	N/A	Full type definitions	N/A

---

Quick Start

Get started with the Burki SDK in your preferred language:

Python

from burki import BurkiClient

# Initialize the client
client = BurkiClient(api_key="your-api-key")

# List all assistants
assistants = client.assistants.list()
for assistant in assistants:
    print(f"{assistant.id}: {assistant.name}")

# Create a new assistant
assistant = client.assistants.create(
    name="Support Bot",
    description="Customer support assistant",
    llm_settings={
        "model": "gpt-4o-mini",
        "temperature": 0.7,
        "system_prompt": "You are a helpful customer support agent."
    },
    tts_settings={
        "provider": "elevenlabs",
        "voice_id": "rachel"
    }
)

---

Real-time Streaming

All SDKs support WebSocket streaming for live transcripts and campaign progress:

Python

import asyncio
from burki import BurkiClient

async def stream_transcripts():
    client = BurkiClient(api_key="your-api-key")
    
    # Stream live transcripts during a call
    async with client.realtime.live_transcript(call_sid="CA123...") as stream:
        async for event in stream:
            print(f"[{event.speaker}]: {event.content}")

asyncio.run(stream_transcripts())

---

Authentication

All SDKs authenticate using your Burki API key. You can generate API keys from your dashboard.

Treat your API keys like passwords! Never share them publicly or commit them to version control.

Environment Variables

For security, we recommend using environment variables:

Python

import os
from burki import BurkiClient

client = BurkiClient(api_key=os.environ["BURKI_API_KEY"])

---

Next Steps

Python SDK

Complete Python SDK documentation with all methods and examples

JavaScript SDK

TypeScript-first SDK with full type definitions

Go SDK

Idiomatic Go SDK with channel-based streaming

Real-time Streaming

WebSocket streaming for live transcripts and campaigns

---

Getting Help

API Reference

Complete REST API documentation

Support

Contact our support team