# Answers API Overview

## Introduction

Welcome to the eGain Answers API Guide. The **Answers API** allows enterprises to deliver fast, accurate, and contextual responses powered by their organizational knowledge. It supports two complementary approaches:

- [**Certified Answers**](/developer-portal/guides/answer/answers_certified): Direct snippets retrieved from enterprise-authored content.
- [**Generative Answers**](/developer-portal/guides/answer/answers_generative): Natural language responses synthesized by a large language model (LLM).


Every response includes supporting search results, references, and confidence scores—ensuring transparency, trust, and traceability. The API is built for secure, scalable integration across enterprise environments.

## What It’s Used For

Organizations use the Answers API to make knowledge more accessible and actionable across different channels. By leveraging both certified and generative answering modes, businesses can:

- Provide consistent, trusted responses across web, mobile, chatbots, and portals.
- Scale customer support without compromising accuracy.
- Empower employees with instant access to critical knowledge.
- Enforce compliance by grounding answers in approved content.


## Common Use Cases

- **Customer Support Automation**: Power chatbots or virtual agents with context-driven answers—reducing costs and improving first-contact resolution.
- **Employee Knowledge Retrieval**: Help employees quickly access policies, technical guides, or internal procedures.
- **Self-Service Portals**: Enable customers to find answers directly in product manuals, help centers, or FAQs.
- **Compliance and Policy Access**: Ensure teams always see approved and up-to-date regulatory or policy information.
- **Research and Analysis**: Surface insights from large collections of reports or reference material.


## Overall Workflow

1. **Query Submitted**
  - A request is sent to the Answers API, optionally including filters such as `topicIds`, `tags`, and `userProfileIds` to narrow the scope.
2. **Certified Answer Retrieval**
  - The API first checks for Certified Answers that meet or exceed the configured `certifiedAnswerThreshold`.
  - The top result above the threshold is returned as the **Certified Answer**, while additional qualifying results are provided in **searchResults**.
3. **Fallback to Generative Answer**
  - If no Certified Answer meets the `certifiedAnswerThreshold`, the query is used to retrieve content using a **hybrid search** retrieval approach.
  - The retrieved chunks along with the query are sent to the LLM. The LLM then generates a natural language response, grounded in retrieved content.
  - The generated answer is checked against a `generativeAnsweTthreshold`. If the confidence score falls below this threshold, a fallback message is returned instead of a potentially unreliable answer.
4. **Supporting Information**
  - All responses include metadata such as references, supporting snippets, and relevance scores to maintain transparency and traceability.


## Next Steps

Take a deeper dive into the capabilities of the Answers API.

- [**Filters**](/developer-portal/guides/answer/answers_filtering): Understand how to narrow results using articleType, topicIds, tags, and user context for precise, personalized answers.
- [**Certified Answers**](/developer-portal/guides/answer/answers_certified): Learn how the API processes queries through Certified and Generative pipelines, and how fallback logic ensures reliable results.
- [**Generative Answers**](/developer-portal/guides/answer/answers_generative): Explore how the API generates natural language responses when certified answers are insufficient, using relevant content chunks and an LLM for contextual, traceable answers.