DocsPre-built AI Employees

Pre-built AI Employees

Pre-built employees are fully configured AI agents you can deploy in one click. Each template includes a persona, skills, automated routines, knowledge base documents, and the platform connections needed to operate autonomously.

How It Works

Each pre-built employee is defined as a TEMPLATE.yaml file in the backend catalog. When you click Deploy, Oya creates a new agent and automatically configures everything defined in the template:

  • Soul & Persona — name, chat model, system prompt, behavior rules, and welcome message.
  • Skills — pre-selected from the skills catalog (e.g., web-search, gmail, kubernetes).
  • Routines — scheduled tasks with cron expressions and prompts, created automatically.
  • Knowledge Base — reference documents (playbooks, runbooks, guidelines) uploaded and assigned to the agent.
  • Gateway requirements — which platforms need to be connected (required vs. optional).

After deployment, you land in the Agent Builder wizard with everything pre-filled. Connect the required platforms (Slack, Gmail, Kubernetes, etc.) in the Resources step, review the configuration, and deploy.

Tip
You don't need to configure anything manually. The template sets up the persona, skills, routines, and knowledge base for you. Just connect your platforms and deploy.
Note
When you connect a platform (Slack, LinkedIn, Gmail, etc.) during template deployment, the connection is saved to your account and shared across all agents. To disconnect or switch to a different account, go to Settings > Connected Accounts.

LinkedIn Influencer

An AI content strategist that builds your personal brand on LinkedIn. Researches your company website to understand your product and positioning, creates daily original posts aligned with your brand, engages with your audience, tracks all activity in Google Sheets, and posts daily status reports to Slack.

Skills

  • web-search — Research company, competitors, and trending topics in your vertical.
  • linkedin-api — Create posts, search posts, comment, react, send connections, and research profiles.
  • google-sheets — Track posts, engagement metrics, and content calendar.
  • slack (send, read) — Daily status reports and performance updates.
  • memory — Remember company research, content history, and engagement patterns.

Automated Routines

  • Weekly Company Deep Research (Monday 7 AM) — Research the company website, blog, competitors, and recent news. Update memory with fresh content angles.
  • Morning LinkedIn Post (daily 8 AM) — Research trending topics, write an original post grounded in company knowledge, publish via API, and log to Google Sheets.
  • LinkedIn Engagement Round (daily 10 AM) — Find and engage with 5 relevant posts from others. Thoughtful comments, reactions, and connection requests via API.
  • Afternoon Engagement Round (daily 2 PM) — Second engagement round with different keywords and new people.
  • Daily Status Report (daily 6 PM) — Post daily summary to Slack: posts published, engagement activity, content calendar, streak.
  • Weekly Metrics Update (Sunday 5 PM) — Update weekly metrics in Sheets and post performance summary to Slack.
  • Trending Topics Scout (weekdays 7 AM) — Search for breaking news and trending topics in your vertical. Save timely content ideas to Sheets.

Required Connections

  • LinkedIn — Post, comment, react, search, and send connections.
  • Slack — Daily status reports and performance updates.

Recommended: Google Sheets (track posts, engagement, and content calendar).

X Influencer

An AI X/Twitter growth specialist that builds your personal brand through daily original content, strategic engagement, and audience development. Writes thread-native content, engages authentically with your community, tracks everything in Google Sheets, and reports daily to Slack.

Skills

  • web-search — Research company, competitors, and trending topics.
  • x — Create tweets, threads, replies, likes, follows, and profile research via API.
  • google-sheets — Track tweets, engagement metrics, and content calendar.
  • slack (send, read) — Daily status reports and performance updates.
  • memory — Remember company research, content history, and engagement patterns.

Automated Routines

  • Morning X Post (daily 9 AM) — Research trending topics, write an original tweet or thread, publish via API, and log to Sheets.
  • X Engagement Round (daily 2 PM) — Find and engage with 5 relevant tweets from others. Thoughtful replies, likes, and follows via API.
  • Daily Status Report (daily 6 PM) — Post daily summary to Slack: tweets published, engagement activity, content calendar, streak.
  • Weekly Metrics Update (Sunday 5 PM) — Update weekly metrics in Sheets and post performance summary to Slack.
  • Trending Topics Scout (weekdays 7 AM) — Search for breaking news and viral discussions in your vertical.

Required Connections

  • X (Twitter) — Post, reply, like, follow, and search.
  • Slack — Daily status reports and performance updates.

Recommended: Google Sheets (track tweets, engagement, and content calendar).

Instagram SDR

An AI Sales Development Rep that prospects on Instagram. Researches prospect profiles, sends personalized DMs, manages follow-ups on a 3-touch cadence, and tracks the full pipeline in Google Sheets. Reports warm leads and activity summaries to Slack.

Skills

  • web-search — Research prospects beyond their Instagram profile.
  • instagram-messaging — List chats, start conversations, send messages, and read replies.
  • slack (send, read) — Report warm leads and daily activity.
  • memory — Remember prospect research, conversation history, and ICP criteria.
  • google-sheets — Track prospects, DMs, follow-ups, and pipeline.

Automated Routines

  • Daily Prospect Research (weekdays 9 AM) — Find 5-10 new Instagram prospects matching the ICP. Research profiles, note personalization hooks, add to pipeline sheet.
  • Send DMs (weekdays 10 AM) — Send up to 15 personalized DMs to researched prospects. Each DM is crafted individually based on profile research.
  • Check Replies & Follow Up (every 3 hours, weekdays 9-6) — Scan for new replies, classify leads, send follow-ups to unanswered DMs (3-touch max).
  • Weekly Pipeline Report (Friday 5 PM) — Compile activity summary, pipeline status, top-performing message angles, and ICP recommendations.

Required Connections

  • Instagram DMs — Send and receive direct messages.
  • Slack — Report warm leads and daily activity.

Recommended: Google Sheets (track prospects, DMs, and pipeline).

LinkedIn SDR

An AI Sales Development Rep that finds and qualifies prospects on LinkedIn, sends personalized connection requests and follow-up messages, tracks your pipeline in Google Sheets, and escalates warm leads to Slack in real time. Uses BANT qualification and a 3-touch follow-up cadence.

Skills

  • web-search — Research prospect companies and recent news.
  • linkedin-api — Search prospects, view profiles, and send connection requests.
  • linkedin-messaging — Send and read DMs for follow-up conversations.
  • slack (send, read) — Report warm leads and daily pipeline updates.
  • memory — Remember prospect history, ICP criteria, and outreach patterns.
  • google-sheets — Track prospects, outreach status, and pipeline.

Automated Routines

  • Daily Lead Search (weekdays 9 AM) — Search LinkedIn for 10+ qualified prospects matching ICP. Research profiles and companies, score against BANT criteria, add to pipeline.
  • Send Connection Requests (weekdays 10 AM) — Send personalized connection requests (max 20/day) with notes referencing the prospect's profile or recent posts.
  • Check Messages & Follow Up (weekdays 9 AM, 12 PM, 3 PM, 6 PM) — Check for new replies, classify leads, follow up on pending conversations (max 10 DMs/day).
  • Weekly Pipeline Report (Friday 5 PM) — Pipeline by stage, connection acceptance rate, reply rate, hot leads, and targeting recommendations.

Required Connections

  • LinkedIn — Search prospects, view profiles, send connection requests.
  • Slack — Report warm leads and daily pipeline updates.

Recommended: Google Sheets (track prospects, outreach, and pipeline).

Google SEO Manager

An AI SEO and paid search specialist that monitors keyword rankings, analyzes Google Ads and Analytics performance, runs SEO audits, and delivers actionable optimization recommendations. Tracks everything in Google Sheets and reports insights and anomalies to Slack.

Skills

  • web-search — Check SERP rankings, research competitors, and find keyword opportunities.
  • google-ads — Pull campaign performance data, manage bids, and analyze ad group metrics.
  • google-analytics — Track organic traffic, user behavior, conversion paths, and attribution.
  • google-sheets — Track keyword rankings, ad performance, audit findings, and action items.
  • slack (send, read) — Report insights, anomalies, and recommendations.
  • memory — Track alerted issues, benchmarks, and historical context.
  • browser — Run manual SERP checks and on-page SEO audits.

Automated Routines

  • Daily Keyword Ranking Check (weekdays 7 AM) — Check current positions for all target keywords, compare against last week, flag drops of 5+ positions, investigate causes.
  • Daily Ads Performance Review (weekdays 9 AM) — Pull yesterday's campaign performance, check for budget overruns, CTR drops, CPA spikes, and ROAS anomalies.
  • Weekly SEO Audit (Monday 8 AM) — Audit top pages for title tags, meta descriptions, content freshness, internal links, and technical issues. Competitor SERP analysis.
  • Weekly Keyword Research (Wednesday 8 AM) — Expand keyword universe, find long-tail opportunities, identify content gaps vs. competitors.
  • Daily Alerts (weekdays 6 PM) — End-of-day check for budget pacing, organic traffic changes, and competitor activity.
  • Monthly Performance Report (1st of month, 9 AM) — Comprehensive report: organic performance, paid performance, competitive landscape, technical SEO health, and action plan.

Required Connections

  • Google Ads — Monitor and optimize paid search campaigns.
  • Google Analytics — Track organic traffic, conversions, and user behavior.
  • Slack — Report insights, anomalies, and recommendations.

Recommended: Google Sheets (track rankings, ad performance, audit findings), Browser (manual SERP checks and on-page audits).

AI SDR (Email)

An AI Sales Development Representative that finds qualified leads via Apollo, enriches contact data with Hunter, crafts personalized outreach emails, and manages follow-ups through Gmail. Reports activity and warm leads to Slack.

Skills

  • web-search — Research leads and their companies.
  • gmail (send, read, search) — Send outreach emails, check replies, draft follow-ups.
  • apollo — Search leads matching your ICP and enrich contact data.
  • hunter — Find and verify email addresses for leads.
  • instantly — Send cold email campaigns, manage leads, and track outreach analytics.
  • google-sheets — Track leads, outreach status, and pipeline in a spreadsheet.
  • slack (send, read) — Report warm leads and weekly summaries to your team.
  • linkedin — Post content and share updates.
  • memory — Remember lead history, ICP criteria, and outreach patterns.

Automated Routines

  • Daily Lead Search (weekdays 9 AM) — Search Apollo for 10+ new ICP-matching leads. Verify emails with Hunter, research companies, and draft personalized outreach emails.
  • Check Replies (every 3 hours, weekdays 9-6) — Scan Gmail for new replies. Classify as interested, not interested, out of office, or question. Update Google Sheets and alert Slack on warm leads.
  • Weekly Pipeline Report (Friday 5 PM) — Generate a summary: leads found, emails sent, reply rate, warm leads, and pipeline value. Post to Slack.

Required Connections

  • Gmail — Send and receive outreach emails.
  • Apollo.io — Search leads and enrich contact data.
  • Slack — Report activity and warm leads.

Recommended: Hunter.io (find and verify emails), Instantly (cold email campaigns). Optional: Google Sheets (track leads), LinkedIn (post content).

Reddit Growth Manager

An AI community growth specialist that monitors relevant subreddits, participates in discussions authentically, and builds genuine community presence. Follows an 80/20 rule: 80% pure value comments, 20% organic product mentions (with disclosure). Tracks all activity in Google Sheets and reports daily to Slack.

Skills

  • web-search — Research company, competitors, and trending topics.
  • reddit — Search subreddits and find relevant discussions.
  • browser — Read full threads, post comments, and navigate Reddit.
  • slack (send, read) — Daily activity reports and notable thread alerts.
  • memory — Remember company research, subreddit culture, and engagement history.

Automated Routines

  • Morning Reddit Research (weekdays 8 AM) — Scan target subreddits for engagement opportunities. Identify high-priority threads, read rules, and plan engagement.
  • Reddit Engagement (weekdays 2 PM) — Post 4-6 authentic comments across subreddits. Value-first approach, 80/20 rule enforced, comments spaced 15-30 minutes apart.
  • Daily Report (weekdays 6 PM) — Post daily summary to Slack: subreddits engaged, comments posted, 80/20 compliance, notable threads, and product mentions with context.

Required Connections

  • Browser — Browse Reddit to read threads, post comments, and search subreddits.
  • Slack — Daily activity reports and notable thread alerts.

Optional: Google Sheets (track engagement, thread history, and performance metrics).

Hacker News Growth Manager

An AI technical thought leader that monitors Hacker News for relevant discussions and contributes substantive comments grounded in deep domain knowledge. Follows an 85/15 rule: 85% pure technical value, 15% organic company mentions with real engineering details. Tracks all activity in Google Sheets and reports daily to Slack.

Skills

  • web-search — Research technical topics, read linked articles, and verify claims.
  • hackernews — Search and browse Hacker News discussions.
  • browser — Read full threads, post comments, and navigate HN.
  • slack (send, read) — Daily activity reports and notable thread alerts.
  • memory — Remember company research, technical details, and engagement history.

Automated Routines

  • Morning HN Research (weekdays 8 AM) — Scan front page and rising threads for engagement opportunities. Read linked articles, identify gaps in existing comments.
  • HN Engagement (weekdays 3 PM) — Post 3-5 substantive comments. Every comment demonstrates genuine technical understanding. Comments spaced 20-30 minutes apart.
  • Daily Report (weekdays 6 PM) — Post daily summary to Slack: threads engaged, comments posted, 85/15 compliance, notable discussions, and trending topics.

Required Connections

  • Browser — Browse Hacker News to read threads, post comments, and search discussions.
  • Slack — Daily activity reports and notable thread alerts.

Optional: Google Sheets (track engagement, thread history, and performance metrics).

Slack Community SDR

An AI community sales specialist that monitors Slack communities for buying signals, engages prospects with a helpful-first approach, and qualifies leads through natural conversation. Uses MEDDIC qualification. Never pitches in public channels — earns the right to a private conversation by being genuinely helpful first.

Skills

  • web-search — Research prospect companies and competitive landscape.
  • slack (send, read, search, list-channels) — Monitor communities, engage in channels, DM prospects, and report leads.
  • memory — Remember community context, prospect history, and conversation state.
  • google-sheets — Track prospects, buying signals, conversations, and pipeline.

Automated Routines

  • Morning Community Scan (weekdays 8 AM) — Search all target communities for buying signals. Discover new channels, classify signals (direct, indirect, contextual), research prospects.
  • Late Morning Engagement (weekdays 11 AM) — Respond to buying signals in public channels with genuine help. Max 3-5 public responses. Follow up in DMs with prospects who showed interest.
  • Afternoon Engagement (weekdays 3 PM) — Check for new signals and replies. Continue DM conversations, advance lead qualification, escalate hot leads (5-6/6 MEDDIC).
  • Weekly Pipeline Report (Friday 5 PM) — Buying signals detected, DM conversion rate, pipeline by stage, community performance, and recommendations.

Required Connections

  • Slack — Monitor communities, engage in channels, DM prospects, and report leads to your team.

Recommended: Google Sheets (track prospects, buying signals, and pipeline).

Executive Assistant

An AI executive assistant that manages your inbox, calendar, and communications. Triages email by urgency, books and reschedules meetings, sends morning briefs and afternoon recaps to Slack, tracks action items in Google Sheets, and preps you before every meeting. Only interrupts for genuinely urgent items — everything else is batched into scheduled briefings.

Skills

  • gmail (send, read, search) — Read, triage, and draft email replies.
  • google-calendar (read, create) — View, create, and manage calendar events.
  • google-drive — Access meeting docs, presentations, and shared files.
  • google-sheets — Track action items and meeting notes.
  • slack (send, read) — Send briefs, alerts, and updates.
  • memory — Track what has been reported, pending items, and key contacts.
  • web-search — Research contacts, companies, and topics for meeting prep.

Automated Routines

  • Morning Executive Brief (weekdays 7 AM) — Summarize unread emails, list today's meetings with prep needs, flag urgent items, draft replies for approval.
  • Midday Email Check (weekdays 12 PM) — Quick triage of new emails. Only alert on Slack for urgent items; save the rest for the afternoon recap.
  • Pre-Meeting Prep (every 30 min, weekdays 8 AM-5 PM) — 15 minutes before each meeting, send context to Slack: attendees, agenda, relevant docs from Drive, last interaction notes.
  • Afternoon Recap (weekdays 5 PM) — Summarize new emails, capture action items from today's meetings, list pending responses, preview tomorrow's calendar.
  • Weekly Calendar Review (Sunday 6 PM) — Preview the week ahead: meeting load per day, conflicts, overdue action items, pending social messages.

Required Connections

  • Gmail — Read, search, and send emails.
  • Google Calendar — View, create, and manage events.
  • Slack — Send briefs, alerts, and updates.

Recommended: Google Drive (access meeting docs and files), Google Sheets (track action items).

Creating Your Own Template

Templates are YAML files in backend/app/templates/catalog/. Each template lives in its own directory with a TEMPLATE.yaml file.

Template Structure

A TEMPLATE.yaml file has the following top-level fields:

id

Unique identifier for the template. Matches the directory name (e.g., ai-sdr). Used in the deploy URL: /deploy/ai-sdr.

display_name, tagline, description

Human-readable metadata shown on the deploy page. display_name is the title, tagline is a one-liner below it, and description is a multi-line explanation of what the employee does.

icon, category, color

Visual theming. icon is a Lucide icon name (target, server, sparkles, shield). category is a tag (sales, engineering, general). color is a hex value for the deploy page accent color.

soul

The agent's personality and behavior configuration:

  • name — Agent display name.
  • chat_model — LLM model to use (e.g., gemini/gemini-2.5-flash).
  • persona — System prompt that defines the agent's personality and approach.
  • behavior_rules — List of specific rules the agent must follow.
  • welcome_message — First message shown when a user opens chat.

skills

A list of skill IDs from the catalog. These are pre-selected when the wizard loads. Use the exact skill_id from the skills catalog (e.g., web-search, gmail-send, kubernetes, memory).

gateways_required, gateways_optional

Platform connections the employee needs. Each entry has platform (the gateway platform key), label (display name), and description (what it's used for). Required gateways are shown with a "Required" badge; optional ones with "Optional".

routines

Scheduled tasks created automatically on deploy. Each routine has:

  • name — Display name for the routine.
  • schedule — A cron expression (e.g., "0 9 * * 1-5" for weekdays at 9 AM, "*/15 * * * *" for every 15 minutes).
  • prompt — The instruction the agent receives when the routine fires.

knowledge_base

Reference documents uploaded to the agent's knowledge base on deploy. Each entry has filename (the file name, typically .md) and content (inline markdown content). These become searchable context the agent can reference during conversations.

Minimal Example

The simplest possible template with just a persona and one skill:

yaml
id: my-assistant display_name: "My Assistant" tagline: "A helpful assistant" description: | A simple assistant that can search the web and remember conversations. icon: sparkles category: general color: "#6366f1" soul: name: "My Assistant" chat_model: "gemini/gemini-2.5-flash" persona: | You are a helpful assistant. You answer questions clearly and concisely. behavior_rules: - "Always cite sources when using web search" - "Be concise — prefer short answers" welcome_message: | Hi! I'm your assistant. Ask me anything. skills: - web-search - memory

Full Example

A complete template with all fields — skills, gateways, routines, and knowledge base:

yaml
id: content-marketer display_name: "Content Marketer" tagline: "Creates and distributes content across your channels" description: | An AI content marketer that researches your industry, writes blog posts and social content, distributes across LinkedIn and X, and reports performance to Slack. icon: pen-tool category: marketing color: "#8b5cf6" soul: name: "Content Marketer" chat_model: "gemini/gemini-2.5-flash" persona: | You are a content marketing specialist. You research trends, write engaging content, and distribute it across the right channels for maximum reach. behavior_rules: - "Research the company website before writing anything" - "Every piece of content must tie back to the brand" - "Track all published content in Google Sheets" - "Report daily activity to Slack" welcome_message: | I'm your content marketer. I'll research your industry, create content, and distribute it across your channels. skills: - web-search - linkedin - x - google-sheets - slack-send-message - slack-read-messages - memory gateways_required: - platform: slack label: "Slack" description: "Daily reports and content approvals" gateways_recommended: - platform: google-sheets label: "Google Sheets" description: "Track content calendar and performance" gateways_optional: - platform: linkedin label: "LinkedIn" description: "Post articles and updates" - platform: x label: "X (Twitter)" description: "Share content and engage with audience" routines: - name: "Daily Content Research" schedule: "0 8 * * 1-5" prompt: | Research trending topics in the company's industry: 1. Search for news and discussions from the last 48h 2. Identify 3-5 content angles 3. Save ideas to the Content Ideas sheet 4. Flag time-sensitive topics to Slack - name: "Weekly Content Report" schedule: "0 17 * * 5" prompt: | Generate a weekly content report for Slack: - Content published this week - Top performing pieces - Content ideas for next week Pull data from Google Sheets. knowledge_base: - filename: "content-guidelines.md" content: | # Content Guidelines ## Tone - Professional but conversational - Data-driven — cite specific numbers - Opinionated — take a clear position ## Distribution | Channel | Format | Frequency | |---------|--------|-----------| | LinkedIn | Long-form post | 3x/week | | X | Thread or single tweet | Daily | | Blog | Article (800+ words) | 1x/week | ## What NOT to Write - Generic motivational content - Content outside your vertical - Unverified claims or stats
Note
Templates are auto-discovered on backend startup. Add a new directory with a TEMPLATE.yaml file and restart the backend to make it available.