AI agent examples

50 AI Agent Examples by Department

Real agents for sales, marketing, support, operations, finance, HR and product — what starts each one, the steps it takes, the tools it uses, and where a person stays in the loop.

An AI agent is software that takes a goal and a set of tools and decides its own steps. Examples: a lead qualifier that scores every inbound lead, a support bot that answers from your help docs, a ticket triager, an SEO auditor, an invoice processor and a pull request reviewer. Below are 50 examples by department, each with its trigger, steps, tools and the point where a person approves.

Sales AI agent examples

1. Inbound lead qualifier

Scores every new inbound lead against your ideal customer profile and routes the good ones to a rep within minutes.

Starts on: Webhook from your form

  1. 1. Reads the form answers and the company's website
  2. 2. Scores fit and intent against your ICP
  3. 3. Updates HubSpot and posts hot leads to Slack

Uses: HubSpot, Web Scraper, Slack

Deploy the Lead Qualification Agent template

2. Lead enrichment table

EV

Fills in missing company size, role and email for every row in a lead list, one agent run per row.

Starts on: New row in a datatable

  1. 1. Looks the person and company up
  2. 2. Finds and verifies a work email
  3. 3. Writes the fields back to the row

Uses: Apollo.io, Hunter.io, Email Verifier

How to build an AI agent for Lead Generation

3. Account research brief

Writes a one-page brief on an account before a call: what they sell, recent news, likely priorities and who to talk to.

Starts on: Calendar event or chat request

  1. 1. Searches the web and recent news
  2. 2. Reads the company's site and LinkedIn page
  3. 3. Writes a brief with sources

Uses: Serper, Web Scraper, LinkedIn, Notion

How to build an AI agent for Sales Research

4. Cold email first-touch writer

Researches each prospect and drafts a first email that references something specific to them.

Starts on: New row in a datatable

  1. 1. Researches the prospect and their company
  2. 2. Drafts a short, specific opener
  3. 3. Queues the draft for review

Uses: Apollo.io, Serper, Instantly

Person approves: Sending — every draft waits for a person

Deploy the Cold Email Outreach Agent template

5. Outreach reply classifier

Reads replies to your campaigns and sorts them into interested, not now, wrong person and unsubscribe.

Starts on: New reply in your sending tool

  1. 1. Reads the reply in context of the thread
  2. 2. Classifies intent and extracts referrals
  3. 3. Updates the lead and alerts the owner

Uses: Instantly, HubSpot, Slack

How to build an AI agent for Email Triage

6. Weekly pipeline report

Every Monday, summarises what moved in the pipeline, what stalled and which deals need attention.

Starts on: Schedule — Mondays 8am

  1. 1. Pulls open deals and stage changes
  2. 2. Flags stalled deals and slipped close dates
  3. 3. Posts a short report to Slack

Uses: HubSpot, Slack

Deploy the Sales Pipeline Reporter template

7. CRM hygiene checker

Finds deals with missing next steps, stale close dates or no owner, and asks the owner to fix them.

Starts on: Schedule — daily

  1. 1. Scans open deals for missing fields
  2. 2. Groups problems by owner
  3. 3. Sends each owner their own fix list

Uses: Salesforce, Slack

How to build an AI agent for Data Entry

8. LinkedIn prospect shortlist

Turns a target-account list into a shortlist of the right people at each company, with a reason for each.

Starts on: New row in a datatable

  1. 1. Finds people matching your target roles
  2. 2. Checks seniority and recent activity
  3. 3. Writes the shortlist back with reasons

Uses: LinkedIn, Apollo.io

How to build an AI agent for LinkedIn Outreach

9. Lead nurture follow-up

Notices leads that went quiet after a demo and drafts a relevant follow-up for the rep.

Starts on: Schedule — daily

  1. 1. Finds leads with no activity for 7 days
  2. 2. Reads the last conversation
  3. 3. Drafts a follow-up for the rep to send

Uses: HubSpot, Gmail

Person approves: Sending — the rep approves each email

How to build an AI agent for Lead Nurturing

10. Local business prospector

Builds a list of local businesses in a category and area, with websites, ratings and a contact route.

Starts on: Chat request

  1. 1. Searches Google Maps for the category and area
  2. 2. Visits each site for contact details
  3. 3. Saves the list to a datatable

Uses: Google Maps Scraper, Web Scraper, Google Sheets

How to build an AI agent for Lead Generation

Marketing AI agent examples

11. SEO site audit

UN
DA

Crawls your site and returns a prioritised list of technical and on-page SEO fixes.

Starts on: Schedule — monthly

  1. 1. Crawls pages and checks titles, links and speed
  2. 2. Compares against search data
  3. 3. Writes a prioritised fix list

Uses: Unlighthouse, DataForSEO

Deploy the Website SEO Audit Agent template

12. Blog research agent

Researches a topic before anyone writes: what ranks, what questions people ask, and sources worth citing.

Starts on: Chat request

  1. 1. Pulls the current top results and questions
  2. 2. Reads the best sources
  3. 3. Writes an outline with citations

Uses: Serper, Firecrawl, Notion

Deploy the Blog Research Agent template

13. Blog draft writer

Turns an approved outline into a full draft in your brand voice, section by section.

Starts on: New row in a content datatable

  1. 1. Expands each section of the outline
  2. 2. Checks facts against the research
  3. 3. Saves the draft for an editor

Uses: Google Docs, Notion

Person approves: Publishing — a person publishes the draft

Deploy the Content Creation Agent template

14. Social post repurposer

Turns each new blog post into a LinkedIn post, an X thread and an Instagram caption.

Starts on: New blog post published

  1. 1. Reads the post and picks the strongest angle
  2. 2. Writes one version per platform
  3. 3. Queues posts for review

Uses: LinkedIn, X (Twitter), Instagram

Person approves: Posting — each post is approved first

Deploy the Social Media Manager template

15. Competitor monitor

WM

Watches competitors' sites and pricing pages and tells you what changed and why it matters.

Starts on: Schedule — weekly

  1. 1. Snapshots competitor pages
  2. 2. Compares them with last week's
  3. 3. Summarises meaningful changes in Slack

Uses: Firecrawl, Wayback Machine, Slack

How to build an AI agent for Competitor Monitoring

16. Keyword gap finder

DA

Finds the searches competitors rank for that you don't, and sorts them by effort and value.

Starts on: Schedule — monthly

  1. 1. Pulls ranked keywords for you and competitors
  2. 2. Finds the gaps
  3. 3. Scores and saves them to a sheet

Uses: DataForSEO, Google Sheets

How to build an AI agent for SEO Research

17. Campaign performance digest

Explains what moved in last week's marketing numbers instead of just listing them.

Starts on: Schedule — Mondays

  1. 1. Pulls traffic, signups and campaign stats
  2. 2. Compares with the previous period
  3. 3. Writes a short digest with likely causes

Uses: Google Analytics, HubSpot Marketing, Slack

How to build an AI agent for Reporting

18. Blog image generator

FA

Creates a header image for every post, in your house style, from the post itself.

Starts on: New blog post draft

  1. 1. Reads the post and picks a visual idea
  2. 2. Generates options
  3. 3. Attaches the best one to the draft

Uses: fal.ai, Stability AI

Deploy the Image Generation Agent template

19. Community thread scout

RE
HA

Finds Reddit and Hacker News threads where people ask about problems you solve, so a person can answer well.

Starts on: Schedule — daily

  1. 1. Searches communities for relevant questions
  2. 2. Filters out threads that would not welcome an answer
  3. 3. Sends a shortlist with context

Uses: Reddit, Hacker News, Slack

Person approves: Replying — a person writes and posts every reply

How to build an AI agent for Social Media Management

20. Newsletter curator

EX

Collects the week's best links in your niche and drafts the newsletter around them.

Starts on: Schedule — weekly

  1. 1. Searches news and community sources
  2. 2. Picks and summarises the best items
  3. 3. Drafts the issue in your sending tool

Uses: Exa.ai, Mailchimp

Person approves: Sending — the issue is approved before it goes out

How to build an AI agent for Content Writing

Customer support AI agent examples

21. Support answer bot

Answers common customer questions from your help docs, and hands anything else to a person with context.

Starts on: New chat or WhatsApp message

  1. 1. Searches your knowledge base
  2. 2. Answers with the relevant doc
  3. 3. Escalates with a summary when unsure

Uses: Intercom, WhatsApp Business

Deploy the Customer Support Bot template

22. Support ticket triager

Tags, prioritises and routes every new ticket so nothing urgent sits in the general queue.

Starts on: New ticket

  1. 1. Reads the ticket and customer history
  2. 2. Sets category, priority and owner
  3. 3. Flags outages and angry customers

Uses: Intercom, Slack

Deploy the Support Ticket Triager template

23. Bug report to Jira

Turns customer bug reports into well-formed Jira issues with steps, impact and linked tickets.

Starts on: Ticket tagged as a bug

  1. 1. Extracts steps to reproduce and impact
  2. 2. Checks for an existing issue
  3. 3. Creates or updates the Jira issue

Uses: Intercom, Jira

How to build an AI agent for Bug Triage

24. Churn risk watcher

Spots accounts showing churn signals — fewer logins, angry tickets, failed payments — and alerts the account owner.

Starts on: Schedule — daily

  1. 1. Combines usage, ticket and billing signals
  2. 2. Scores each account's risk
  3. 3. Alerts owners with the reasons

Uses: Mixpanel, Intercom, Stripe, Slack

How to build an AI agent for Churn Prevention

25. NPS comment analyst

Reads every survey comment and turns them into themes with counts and quotes.

Starts on: Schedule — monthly

  1. 1. Collects survey responses
  2. 2. Groups comments into themes
  3. 3. Writes a report with representative quotes

Uses: Google Sheets, Notion

How to build an AI agent for NPS Surveys

26. Customer onboarding guide

Checks how far each new customer got through setup and sends the right next step.

Starts on: Schedule — daily

  1. 1. Reads each new account's setup progress
  2. 2. Picks the next step that would help
  3. 3. Drafts a short personal email

Uses: Segment, SendGrid

Person approves: Sending — emails are approved in batches

How to build an AI agent for Customer Onboarding

27. WhatsApp order intake

Takes orders and enquiries over WhatsApp, asks for missing details, and logs each one in a table.

Starts on: New WhatsApp message

  1. 1. Understands what the customer wants
  2. 2. Asks for anything missing
  3. 3. Logs the order and confirms it

Uses: WhatsApp Business, Google Sheets

How to build an AI agent for Order Processing

28. Product feedback router

Collects feature requests from tickets and calls and files them against the right roadmap item.

Starts on: New ticket or call note

  1. 1. Finds feature requests in the text
  2. 2. Matches them to existing requests
  3. 3. Adds a vote and quote to the roadmap

Uses: Intercom, Notion

How to build an AI agent for Feedback Collection

Operations AI agent examples

29. Meeting notes summariser

Turns meeting notes into decisions, owners and deadlines, and posts them where the team works.

Starts on: New notes in Notion

  1. 1. Reads the notes
  2. 2. Extracts decisions and action items
  3. 3. Posts a summary and creates tasks

Uses: Notion, Slack, Asana

Deploy the Meeting Notes Summarizer template

30. Shared inbox triage

Sorts a shared inbox into needs-reply, FYI and spam, and drafts replies to the routine ones.

Starts on: New email

  1. 1. Reads and classifies each email
  2. 2. Drafts replies to routine requests
  3. 3. Labels and assigns the rest

Uses: Gmail, Slack

Person approves: Replying — drafts wait for a person

How to build an AI agent for Email Triage

31. Document data extractor

Pulls the fields you need out of PDFs and emails and enters them into your system.

Starts on: New file in Google Drive

  1. 1. Reads the document
  2. 2. Extracts and validates the fields
  3. 3. Writes them to Airtable and flags doubts

Uses: Google Drive, Airtable

How to build an AI agent for Data Entry

32. Monthly delivery report

Writes the monthly report of what the team shipped, what slipped and why, from Jira.

Starts on: Schedule — monthly

  1. 1. Pulls completed and overdue issues
  2. 2. Groups them by project
  3. 3. Writes the report as a Google Doc

Uses: Jira, Google Docs

Deploy the Jira Monthly Report Generator template

33. Vendor shortlist researcher

Researches vendors for a need you describe and returns a comparison with pricing and caveats.

Starts on: Chat request

  1. 1. Searches for vendors that fit
  2. 2. Reads their pricing and review pages
  3. 3. Writes a comparison table with sources

Uses: Tavily, Web Scraper, Notion

How to build an AI agent for Vendor Management

34. Inventory reorder planner

Watches stock levels and sales pace, and proposes reorders before you run out.

Starts on: Schedule — daily

  1. 1. Reads stock and recent sales
  2. 2. Forecasts days of cover
  3. 3. Proposes reorders to the inbox

Uses: Google Sheets, Slack

Person approves: Ordering — a person accepts each proposal

How to build an AI agent for Inventory Management

35. Overdue task chaser

Finds overdue tasks, works out what is blocking them, and nudges the right person.

Starts on: Schedule — daily

  1. 1. Finds overdue tasks
  2. 2. Reads comments for blockers
  3. 3. Sends each owner a specific nudge

Uses: Asana, Trello, Slack

How to build an AI agent for Reporting

36. Custom API runner

Calls an internal or third-party API on your behalf and explains the result in plain English.

Starts on: Chat request or webhook

  1. 1. Works out which endpoint to call
  2. 2. Calls it with the right parameters
  3. 3. Summarises the response

Uses: HTTP Request, Webhook

Person approves: Writes — any change-making call waits for approval

Deploy the API Caller Agent template

Finance AI agent examples

37. Invoice processor

Reads incoming invoices, matches them to vendors and purchase orders, and prepares them for payment.

Starts on: New invoice email

  1. 1. Extracts invoice details
  2. 2. Matches vendor and amount
  3. 3. Creates a draft bill for approval

Uses: Gmail, QuickBooks

Person approves: Paying — bills are approved by finance

How to build an AI agent for Invoice Processing

38. Expense categoriser

Categorises expenses and flags the ones that break policy before month end.

Starts on: New expense

  1. 1. Reads the receipt and description
  2. 2. Assigns a category and checks policy
  3. 3. Flags exceptions to finance

Uses: Xero, Slack

How to build an AI agent for Expense Tracking

39. Failed payment recovery

Follows up on failed card payments with a helpful, personal message instead of a generic dunning email.

Starts on: Failed payment event

  1. 1. Reads the customer's history
  2. 2. Drafts a message with the fix link
  3. 3. Schedules a follow-up

Uses: Stripe, SendGrid

Person approves: Sending — messages are approved first

How to build an AI agent for Payment Recovery

40. Bookkeeping reconciler

Matches bank transactions to invoices and bills and lists what it could not match.

Starts on: Schedule — weekly

  1. 1. Pulls new transactions
  2. 2. Matches them to open items
  3. 3. Lists unmatched items with suggestions

Uses: QuickBooks, Google Sheets

How to build an AI agent for Bookkeeping

41. Revenue digest

Explains the week's revenue: new, expansion, churn, and the customers behind each number.

Starts on: Schedule — Mondays

  1. 1. Pulls payments and subscription changes
  2. 2. Groups them into movements
  3. 3. Writes a short digest

Uses: Stripe, Slack

How to build an AI agent for Reporting

People & HR AI agent examples

42. Candidate screener

Reads applications against the role's must-haves and shortlists with a reason for each decision.

Starts on: New application

  1. 1. Reads the CV and answers
  2. 2. Checks them against the must-haves
  3. 3. Shortlists with reasons for a recruiter

Uses: Gmail, Airtable

Person approves: Rejecting — a person makes every rejection

How to build an AI agent for Recruiting

43. New hire onboarding buddy

Answers a new hire's questions from your handbook and checks their first-week tasks are done.

Starts on: Slack message

  1. 1. Searches the handbook
  2. 2. Answers or points to the right person
  3. 3. Tracks onboarding tasks

Uses: Slack, Notion

How to build an AI agent for Customer Onboarding

44. Performance review prep

Gathers a person's shipped work and feedback over the period so the review starts from facts.

Starts on: Chat request

  1. 1. Collects completed work and notes
  2. 2. Groups it by goal
  3. 3. Drafts a summary for the manager

Uses: Jira, GitHub, Notion

How to build an AI agent for Performance Reviews

45. Async standup collector

Collects daily updates, spots blockers, and posts one summary instead of twenty messages.

Starts on: Schedule — daily

  1. 1. Asks each person for their update
  2. 2. Spots blockers and overlaps
  3. 3. Posts one summary

Uses: Microsoft Teams, Slack

How to build an AI agent for Meeting Notes

Product & engineering AI agent examples

46. Pull request reviewer

Reviews each pull request for bugs, risky changes and missing tests before a human looks.

Starts on: New pull request

  1. 1. Reads the diff and linked issue
  2. 2. Comments on bugs and risks
  3. 3. Summarises what to check

Uses: GitHub

Deploy the PR Code Reviewer template

47. Site performance audit

UN

Runs performance checks across your key pages and explains which fixes matter most.

Starts on: Schedule — weekly

  1. 1. Runs Lighthouse across pages
  2. 2. Compares with last run
  3. 3. Writes a prioritised fix list

Uses: Unlighthouse, Slack

Deploy the Site Performance Audit Agent template

48. Bug triage

Deduplicates incoming bug reports, estimates severity, and assigns them to the right team.

Starts on: New issue

  1. 1. Checks for duplicates
  2. 2. Estimates severity and area
  3. 3. Labels and assigns the issue

Uses: Jira, GitHub

How to build an AI agent for Bug Triage

49. Changelog writer

Turns merged pull requests into a customer-readable changelog entry.

Starts on: Schedule — weekly

  1. 1. Reads merged PRs and linked issues
  2. 2. Picks the changes customers will notice
  3. 3. Drafts the changelog entry

Uses: GitHub, Notion

Person approves: Publishing — a person publishes the entry

How to build an AI agent for Changelog

50. User research synthesiser

Turns interview notes into themes, pains and quotes your team can act on.

Starts on: New notes in Notion

  1. 1. Reads interview notes
  2. 2. Groups findings into themes
  3. 3. Writes a synthesis with quotes

Uses: Notion

How to build an AI agent for User Research

AI agent examples: common questions

An AI agent is software that is given a goal and a set of tools, and decides for itself which steps to take to reach the goal. Where a rule-based automation follows the same fixed steps every time, an agent reads the situation — an email, a lead, a ticket — and chooses what to do next.
Common examples include lead qualifiers that score inbound leads, support bots that answer from your help docs, ticket triagers, SEO auditors, invoice processors, candidate screeners and pull request reviewers. This page lists 50, grouped by department, each with its trigger, steps, tools and the point where a person approves.
An automation runs the same steps every time ("when a form is submitted, add a row"). An agent decides the steps on each run ("read this lead, research the company, decide whether it fits, then route it"). Use an automation when every case is the same; use an agent when each case needs judgement.

Build one of these without code

Start from a template in the agent library, or describe the job and let the agent creator set it up.

Last reviewed 2026-10-01.