Skip to content

Next Steps

Congratulations on completing the Getting Started guides! You now have a working Pika installation and understand the basics of creating agents. This page will help you decide where to go next based on your goals.

If you're ready to build real applications for your users:

Implement Real Authentication

Replace the mock authentication provider with a real one. This is critical before deploying to production.

Authentication Guide →

Configure Access Control

Set up user types and chat app visibility to control who can access what.

Access Control Guide →

Set Up CI/CD

Automate your deployment pipeline with continuous integration and deployment.

CI/CD Guide →

If you want to deepen your understanding of how Pika works:

Explore Capabilities

Discover all the features and capabilities Pika offers for building intelligent chat applications.

Capabilities Overview →

Understand the Architecture

Learn how Pika works under the hood, including AWS infrastructure, session management, and the self-correcting loop.

Concepts & Architecture →

Read the Design Philosophy

Understand why Pika was built and the problems it solves.

Why Pika →

If you're ready to add sophisticated capabilities to your agents:

Add Tools to Your Agents

Learn how to give your agents access to external APIs, databases, and services using inline tools.

Inline Tools Guide →

Integrate with MCP

Connect to Model Context Protocol servers to give your agents access to external tools and data sources.

MCP Integration Guide →

Enable Multi-Agent Collaboration

Create systems where multiple specialized agents work together to solve complex problems.

Multi-Agent Guide →

Build Custom Web Components

Create custom UI components to render rich, interactive content in chat responses.

Web Components Guide →

If you want to make your agents more reliable and accurate:

Enable Self-Correcting Responses

Use Pika's self-correcting loop to automatically verify and improve agent responses.

Self-Correcting Guide →

Set Up LLM Feedback

Enable AI-generated feedback on chat sessions to identify improvement opportunities.

LLM Feedback Guide →

Master Instruction Engineering

Learn how to write effective instructions and use Pika's instruction assistance tools.

Instruction Engineering →

If you want to tailor Pika to your brand and requirements:

Configure User Memory

Set up persistent user memory so agents remember context across sessions.

User Memory Guide →

  1. Why Pika - Understand the value proposition
  2. Who Is Pika For - See if it fits your use case
  3. Capabilities Overview - Learn what you can build
  4. Pika vs Other Solutions - Compare alternatives
  1. What Are AI Agents - Understand the basics
  2. How Pika Works - Learn the architecture
  3. Agent Development Basics - Start building
  4. Tools and Integration - Add capabilities
  1. Agent-as-Config Philosophy - Understand the approach
  2. Multi-Agent Orchestration - Advanced patterns
  3. Self-Correcting Loop - Quality improvements
  4. Direct Agent Invocation - Programmatic access
  1. AWS Architecture - Understand infrastructure
  2. AWS Deployment - Deploy to production
  3. Security Architecture - Security model
  4. Monitoring and Traces - Operational tooling
  1. Identify a specific use case (customer support, internal knowledge base, etc.)
  2. Design your agent's capabilities and knowledge requirements
  3. Implement necessary tools for data access
  4. Configure authentication and access control
  5. Test thoroughly with real users
  6. Deploy to production with monitoring
  1. Map your existing authentication to Pika's user types
  2. Create tools to access your databases and APIs
  3. Build custom web components for your data types
  4. Configure access control based on your org structure
  5. Embed Pika chat into your existing applications
  1. Identify specialized domains (e.g., sales, support, technical)
  2. Create agents for each domain with specific instructions
  3. Configure agent collaboration patterns
  4. Test coordination between agents
  5. Monitor performance and iterate

Issue Tracker

Report bugs, request features, or check on known issues.

GitHub Issues →

Keep up with Pika developments:

If you're stuck or have questions:

  1. Check the docs - Search for your specific topic
  2. Review existing issues - Someone may have encountered the same problem
  3. Ask in Discussions - The community is here to help
  4. Create an issue - Report bugs or request clarification

The best way to learn Pika is by building with it. Start with a simple use case, deploy it, get feedback, and iterate. The Pika community is excited to see what you build!

Ready to build something amazing? Pick a path above and get started!