Cody
Train Cody on your company docs, policies, and processes. When customers ask questions, your Tars AI agent queries Cody's trained bots, retrieves sourced answers from your knowledge base, and responds with accurate, citation-backed information. Support costs drop while answer quality climbs.
Your AI agent taps into Cody's trained bots and document library to pull verified answers, manage conversations, and keep your knowledge base current in real time.
Cody
See how businesses use Tars and Cody together to transform scattered internal knowledge into instant, accurate answers for every customer interaction.
A customer asks about your return policy at midnight. Your AI Agent sends the question to Cody's support bot, which is trained on your latest policy documents. Cody returns the exact return window, conditions, and process steps with source citations. The customer gets a precise answer in seconds, and your team avoids another routine policy ticket in the morning queue.
A user reports an error message in your software. Your AI Agent creates a focused Cody conversation scoped to your troubleshooting guides and known-issues database. Cody identifies the matching resolution steps, and the agent walks the customer through the fix. Complex issues that once required Level 2 support now resolve in a single chat session.
A new customer asks how to set up your product. Your AI Agent queries the Cody bot trained on getting-started docs and tutorial content. It retrieves step-by-step setup instructions, configuration options, and best practices. The customer completes onboarding independently, and your success team focuses on strategic accounts instead of repetitive walkthroughs.
Cody
FAQs
The agent uses Cody's List Bots endpoint to see all available bots and their trained domains. Based on the customer's question topic, it routes to the appropriate specialist bot. A product question goes to your product-trained bot, while a billing question goes to your finance-trained bot. You can also configure default routing rules in your agent settings.
Yes. When creating a conversation, the agent can pass document_ids to enable focus mode. This restricts Cody's answers to only those specific files. Useful when a customer asks about a particular product version or policy document, ensuring the response references only the relevant content rather than your entire knowledge base.
Yes. Cody includes sources with every response, showing which documents and passages informed the answer. Your AI agent can optionally share these citations with the customer, adding credibility. Include the 'sources' parameter when retrieving messages to get the full citation details alongside each response.
When Cody cannot find a relevant answer in its trained documents, it indicates low confidence. Your Tars agent detects this and can escalate to a human agent, offer to create a support ticket, or ask the customer to rephrase. No fabricated answers, just transparent handoffs when knowledge gaps exist.
Yes. The integration supports creating documents from text content, uploaded files up to 100 MB, or public webpage URLs. Your team can feed new product updates, policy changes, or FAQ articles into Cody directly through Tars workflows. Supported formats include PDF, DOCX, PPTX, Markdown, and plain text.
Cody's native widget provides knowledge base Q&A. Tars adds multi-channel deployment across website, WhatsApp, and SMS, plus the ability to chain Cody queries with actions in other tools like your CRM or ticketing system. A single conversation can answer a knowledge question and then create a Zendesk ticket if needed.
Tars passes messages to Cody in real time and receives responses during the conversation. Message history is maintained in Cody's conversation threads via their API. Tars does not maintain a separate copy of your Cody conversation data. Your knowledge base content stays within Cody's AWS-encrypted infrastructure.
Yes. Cody's API supports focus mode with up to 1,000 document IDs per conversation. The agent can scope large document sets for complex queries spanning multiple manuals or policy collections. For most customer questions, a narrower focus of 5 to 20 documents produces faster, more precise answers.
Don't limit your AI Agent to basic conversations. Watch how to configure and add powerful tools making your agent smarter and more functional.
Privacy & Security
At Tars, we take privacy and security very seriously. We are compliant with GDPR, ISO, SOC 2, and HIPAA.
Email - sales@hellotars.com
.png)
%201.png)
%201.png)
