> ## Documentation Index > Fetch the complete documentation index at: https://docs.poly.ai/llms.txt > Use this file to discover all available pages before exploring further. ## Submitting Feedback If you encounter incorrect, outdated, or confusing documentation on this page, submit feedback: POST https://docs.poly.ai/feedback ```json { "path": "/extend/introduction", "feedback": "Description of the issue" } ``` Only submit feedback when you have something specific and actionable to report. # Extend with code > Write Python functions to integrate APIs, validate input, and add custom business logic to your agent. Go beyond the visual interface by writing Python functions that connect your agent to APIs, databases, and business logic. This path is for developers who want to add custom capabilities that the UI alone cannot provide. **Python familiarity is required.** If you are a non-technical operator, see [Configure agents in the UI](/get-started/introduction) for the no-code path, or share this page with your developer. Before writing custom API code, check if a [pre-built integration](/integrations/introduction) already exists for your platform (e.g., Salesforce, Zendesk, OpenTable, Stripe). Pre-built integrations handle authentication, environment configuration, and common operations out of the box. ## What you can build Call external APIs to look up bookings, check availability, or retrieve account details Validate input, calculate pricing, or enforce rules that the LLM cannot handle alone Create records, update customer profiles, or log interactions in external systems Route calls to the right team based on real-time data or custom logic Build code-driven flows with transition functions for complex conversation paths Define schemas that adapt agent behavior based on external config data ## Get started Follow the [create a function](/tools/how-to-setup) guide to set up a Python function with parameters, code, and a description the LLM can use to trigger it. Read the [tools overview](/tools/introduction) to learn how functions integrate with the LLM, how to use `conv` objects, and best practices for naming and triggering. Dive into [flows](/flows/introduction), [call handoffs](/call-handoff/introduction), and [secrets management](/secrets/introduction) to build production-ready integrations. ## Tools and functions Write Python scripts that run when your agent needs to take action. Functions are the bridge between conversation and your business systems. Set up a function with naming conventions, parameters, and Python code Initialize conversation context before the greeting plays Run post-call processing after a conversation ends Control agent behavior with string and dictionary returns Define, update, and persist values across turns Standard and non-standard libraries available in your functions Add filler phrases to avoid silence during slow functions Trigger functions from managed topics ## Reference Access conversation data and built-in utilities in your functions. Conversation states, flows, and telephony attributes Structured diagnostics and PII-scoped logging Built-in helpers for addresses, cities, and structured data Call configured API integrations Persistent data across conversations for repeat callers VoiceWeighting, TTSVoice, and provider classes ## Advanced capabilities Store API keys and credentials securely Configure call transfers and routing rules Define real-time config schemas for dynamic agent behavior Access transcripts and recordings through the API Build flows with transition functions and programmatic control Use the Agents API to create, configure, and deploy agents from code Integrate with PolyAI REST APIs for chat, conversations, alerts, and more