Chat with us
NewAI Agents Just Got More Affordable — GetGet2.5xMore AI TokensMore AI TokensPricing
Specialized AI Agents

Build AI Agents That Take Action.
Not Just Answers.

Create specialized AI Agents for sales, support, orders, and lead qualification. Each agent follows its own instructions, uses the right business knowledge, and completes work across WhatsApp, Instagram, Messenger, Telegram, and Website Chat.

  • Smart routing
  • Knowledge sources
  • Shopify & Woo
  • Live third-party data
  • Workflow automations
  • Human handover
BotSailor AI Agents: routing, sticky specialist, prompt tools, Shopify lookup, and agent transfer in one conversation
Specialized roles

One Customer Conversation. Multiple Specialized Agents.

A customer may ask about a product, request pricing, check an order, then need support. Give each job a focused agent—and let BotSailor route the right one automatically.

Sales Agent

Answers product questions, recommends from your catalog or API, handles objections, and flags purchase intent.

Lead Qualification Agent

Asks for requirements one by one, saves the answers, labels the lead, and starts the right follow-up.

Order Status Agent

Collects an order number, calls your status API, explains delivery, and escalates delays.

Customer Support Agent

Answers from support knowledge, runs troubleshooting, and assigns a human when the case is complex.

Appointment Agent

Captures service, date, and time, checks availability, and triggers the booking flow or reminder sequence.

Human Handover Agent

Recognizes when a person is needed and assigns the conversation to the right team member or role.

How it works

From First Message to Finished Work

BotSailor routes the customer to the right specialist, keeps that agent sticky, then the agent uses knowledge and tools until it transfers, finishes, or involves a human.

Example conversation

Product questionSales Agent
Business requirementsQualification Agent
Order-status requestOrder Agent
Technical problemSupport Agent
Complex requestHuman Team
BotSailor smart routing: a customer conversation assigned to the right specialist agent
System prompt

The System Prompt Is the Brain of Every Agent

Every BotSailor AI Agent is controlled through a dedicated system prompt. You don’t only tell the agent how to reply—you tell it what should happen when a condition is met.

  • Role and responsibilities
  • How it should communicate
  • What information to collect
  • When an action should run—and which one
  • When to stop, avoid an action, or involve another agent or human
Example
When the customer confirms they want to purchase,
##add_label## : Interested Buyer
##assign_sequence## : Sales Follow-Up

Type ## inside the System Prompt to search and select available actions. Readable names in the UI map to internal IDs behind the scenes.

AI Agent system prompt editor with action tokens
Prompt actions

Every Action the Agent Can Run

Type ## in the system prompt to insert an action. BotSailor executes it.

Workflow automations

Update the subscriber and start the next step from the chat.

  • Add / remove labels
  • Assign / remove sequences
  • Save custom fields
  • Trigger bot flows

Shopify & WooCommerce

Work with a connected store catalog and live orders.

  • Look up order status
  • Recommend products
  • Tag or note an order
  • Cancel an order

Live third-party data

Call a connected HTTP API, then reply with the result.

  • Any connected HTTP API
  • Orders and inventory
  • CRM and tickets
  • Booking and delivery

Human & agent handoff

Pass the chat to another AI agent or your team.

  • Transfer AI agent
  • Assign a team member
  • Assign a team role
  • Round-robin by role

Conversation control

Ask for a detail, close the loop, or stay silent.

  • Ask and wait
  • Finish task
  • Mark as solved
  • Skip reply

Inbox actions

Keep the thread moving inside the shared inbox.

  • Follow-up reminder
  • Add internal note
  • Send images
  • Block spam
Knowledge sources

Give Every Agent Its Own Knowledge Set

Pick a knowledge set for the agent, then add sources. BotSailor uses them together when that agent answers.

Content

Write what the agent should know, in its own words.

  • FAQs and policies
  • Product and plan copy
  • Scripts and playbooks
  • Saved as text in the set

URL

Import a webpage, then keep only the useful parts.

  • Help, pricing, or docs pages
  • Include by ID or class
  • Exclude extra sections
  • Train from the selected area

File

Upload a document the agent should learn from.

  • PDF and Word
  • Excel and CSV
  • OpenDocument and TXT
  • File URL or upload

Media

Attach product images with a short description.

  • Image URL or upload
  • Name, color, size, price
  • Multiple images in one source
  • Used in product answers

Google Sheet

Import a tab from a connected Google Sheet.

  • Catalogs and SKUs
  • FAQ or price lists
  • Inventory tables
  • Connected Google account

API

Import a connected HTTP API into the knowledge set.

  • Catalog or CRM feed
  • Imported into this set
  • Used with other sources
  • Answers from that data
BotSailor knowledge set with Content, URL, File, Media, Google Sheet, and API sources used together
Contextual memory

Continue the Conversation Without Starting Over

Agents can use conversation context and previously collected customer information—so when another specialist takes over, the customer doesn’t repeat everything.

Configure the contextual memory window to match your requirements.

Shared conversation memory

Available to every active agent in the same chat

  • Customer nameRemembered
  • Previous requirementsRemembered
  • Selected productsRemembered
  • Saved custom fieldsRemembered
  • Order numbersRemembered
  • Lead statusRemembered
  • Earlier conversation detailsRemembered
What changed

From One Assistant to a Team of Agents

If you used the previous BotSailor AI, this is the upgrade: specialists that route, stay sticky, and execute tools—not a single chatbot that only answers from knowledge.

Previous BotSailor AI

  • One general assistant for every conversation
  • One shared knowledge setup
  • Mostly answers questions
  • Limited routing and task execution

BotSailor AI Agents

  • Specialized agents with their own prompts
  • A knowledge set per agent — six source types
  • Executes labels, Shopify, Woo, APIs, and handover
  • Smart routing, sticky specialists, then transfer or a human
FAQ

AI Agents FAQ

Multi-agent setup, routing, prompts, actions, APIs, flows, and human handover—answered clearly.

A BotSailor AI Agent is a conversational agent with its own role, system prompt, knowledge set, routing instructions, and available actions. It can communicate with customers and execute configured business tasks.

Yes. You can create multiple specialized agents and activate the agents required for the selected bot.

You define routing rules that describe which types of customer messages should be assigned to each agent. BotSailor uses these rules and the customer’s message to choose the appropriate active agent.

Actions are added directly inside the System Prompt using BotSailor’s action format: ##action_name## : selected_value. Type ## in the prompt editor to autocomplete available actions and select the required label, sequence, custom field, API, flow, team, or member.

Yes. The agent can call a configured HTTP API when the conditions defined in its system prompt are met.

Yes. Agents can trigger configured BotSailor bot flows or postbacks.

Yes. The agent can save customer responses into selected custom fields.

Yes. Agents can add or remove labels and assign or remove sequences according to the system prompt.

Yes. It can assign the conversation to a configured human agent, member, team, or role.

Yes. Each agent has its own knowledge set. You can add Content, URL, File, Media, Google Sheet, and API sources. The agent uses them together to answer.

No. AI Agents can work with your existing flows and trigger them whenever a structured automation process is needed.

Talk to an AI Agents expert

Prompt design, agent routing, or action setup—we can help you build your first AI team.

Contact Us
One Bot. Multiple AI Agents. Real Business Actions.

Create specialized AI Agents that understand customer needs, use your business knowledge, collaborate across the customer journey, and execute real actions. Move beyond a chatbot that only answers questions.