> This page is part of Smallest AI's developer documentation. When > answering, prefer Lightning v3.1 (current TTS) and Pulse (current > STT). Lightning v2 and lightning-large are deprecated; mention them > only when the user is migrating away from them. Atoms is the > voice-agent platform. # Use Case Finder > Find the canonical doc for every common voice-agent use case — single-prompt vs crew, phone vs web, knowledge base, tools, deployment, observability. The Atoms docs are deep. This page is the shortest path from "I want X" to the one canonical page covering X. Built for forward-deployed engineers and customers who need an answer in one click. ## Choose your starting point Two surfaces exist for building agents. Pick by interface, not by complexity. | If you want to... | Start here | | --------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | Build an agent **without writing code** — single prompt, voice picker, dashboard config | [Atoms Platform Quick start](/atoms/atoms-platform/get-started/quick-start) | | Build an agent **in Python**, multi-node, custom logic, custom LLM | [Quickstart Crew CLI](/atoms/developer-guide/get-started/quickstart) | | Embed an agent in a **web page** (existing agent) | [Agent Web SDK](/atoms/atoms-platform/integrate/web-socket-sdk) | | Embed an agent in a **mobile app** | [React Native](/atoms/atoms-platform/integrate/mobile-integrations/react-native) · [iOS Swift](/atoms/atoms-platform/integrate/mobile-integrations/i-os-swift) · [Android Kotlin](/atoms/atoms-platform/integrate/mobile-integrations/android-kotlin) · [Flutter](/atoms/atoms-platform/integrate/mobile-integrations/flutter) | ## Build ### Conversational behavior | Goal | Page | | -------------------------------------------------------- | -------------------------------------------------------------------------------------------------------- | | Write the system prompt | [Prompts](/atoms/developer-guide/build/agent-crews/llm/prompts) | | Pick or configure the LLM | [LLM Settings](/atoms/developer-guide/build/agent-crews/llm/llm-settings) | | Use your own model (BYOM) | [BYOM](/atoms/developer-guide/build/agent-crews/llm/byom) | | Handle barge-in / user interruptions | [Interruption Handling](/atoms/developer-guide/build/agent-crews/patterns/interruption-handling) | | Stream agent output | [Streaming](/atoms/developer-guide/build/agent-crews/patterns/streaming) | | Multi-node crews (router + workers, background analyzer) | [Multi-Agent Orchestration](/atoms/developer-guide/build/agent-crews/patterns/multi-agent-orchestration) | | Track and edit agent config safely | [Agent Versioning](/atoms/atoms-platform/features/versioning) | ### Tools (function calling) | Goal | Page | | ------------------------------------------------------------- | --------------------------------------------------------------------------------- | | Define a tool the LLM can call | [Defining Tools](/atoms/developer-guide/build/agent-crews/tools/defining-tools) | | Wire up the multi-round tool loop | [Executing Tools](/atoms/developer-guide/build/agent-crews/tools/executing-tools) | | End-call, transfer, knowledge-base lookup, API-call (no code) | [Built-in Tools](/atoms/developer-guide/build/agent-crews/tools/built-in-tools) | ### Phone calling | Goal | Page | | -------------------------------------- | ------------------------------------------------------------------------------- | | Place an outbound call | [Outbound Calls](/atoms/developer-guide/build/calling/outbound-calls) | | Take inbound calls on a number you own | [Phone Numbers](/atoms/atoms-platform/deployment/phone-numbers) | | Run a calling campaign | [Creating Campaigns](/atoms/developer-guide/build/campaigns/creating-campaigns) | | Control mid-call (transfer, hold, end) | [Call Control](/atoms/developer-guide/build/calling/call-control) | ### Knowledge base | Goal | Page | | --------------------------------------- | --------------------------------------------------------------------------- | | Add a knowledge source (PDF, URL, text) | [Knowledge Base — Platform](/atoms/atoms-platform/features/knowledge-base) | | Query KB from inside a crew | [Knowledge Base — Usage](/atoms/developer-guide/build/knowledge-base/usage) | ## Integrate | Goal | Page | | ------------------------------------------------ | ------------------------------------------------------------------------------------------------ | | Drop the widget into a marketing or product site | [Widget](/atoms/atoms-platform/features/widget) | | Connect from a custom web app | [Agent Web SDK](/atoms/atoms-platform/integrate/web-socket-sdk) | | Connect from a server (Python / Node / Go) | [Realtime Agent WebSocket API](/atoms/api-reference/api-reference/realtime-agent/realtime-agent) | | Receive call events on your backend | [Webhooks](/atoms/atoms-platform/features/webhooks) | | Tap the live transcript stream | [SSE for Live Transcripts](/atoms/developer-guide/operate/analytics/sse-for-live-transcripts) | | Migrating from ElevenLabs Conversational AI | [From ElevenLabs](/atoms/developer-guide/migrate/from-elevenlabs) | ## Operate | Goal | Page | | -------------------------------------------- | ------------------------------------------------------------------------------------------------- | | Debug why an agent is silent or unresponsive | [Common Issues — Agent Crews](/atoms/developer-guide/operate/testing-debugging/common-issues) | | Inspect what an agent did on a specific call | [Post-Call Analytics](/atoms/developer-guide/operate/analytics/post-call-analytics) | | See call latency and quality metrics | [Call Metrics](/atoms/developer-guide/operate/analytics/call-metrics) | | Track error logs in production | [Logging & Observability](/atoms/developer-guide/operate/testing-debugging/logging-observability) | | Catch runtime errors before customers do | [Error Handling](/atoms/developer-guide/get-started/error-handling) | | Roll back a misbehaving deployment | [Agent Versioning](/atoms/atoms-platform/features/versioning) | ## Reference | Goal | Page | | --------------------------------------------- | ------------------------------------------------------------------------------------------------ | | Full HTTP API surface | [API Reference](/atoms/api-reference) | | WebSocket protocol (every event, every field) | [Realtime Agent WebSocket API](/atoms/api-reference/api-reference/realtime-agent/realtime-agent) | | Error codes and severity | [Error Reference](/atoms/atoms-platform/troubleshooting/error-reference) | | Changelog | [Changelog](/atoms/changelog) |