Building PayCaptain's Open Employee Handbook with Lovable

What they built

• A public-facing, branded employee handbook for PayCaptain (47-person remote UK scale-up), built in Lovable instead of Notion or PDF
• Covers policies (maternity/parental leave, menopause, compensation philosophy, levelling frameworks, referral policy, and more), team info, and a community section for customers and partners
• Includes an AI chat assistant embedded in the handbook that answers questions using only handbook content - with a guardrail preventing fabrication, defaulting to "contact Emily" if information isn't found
• Connected to a Slack bot (built via n8n) so employees and managers can query handbook content without leaving Slack
• Integrated with Granola (AI meeting transcription) → Slack → n8n → Lovable pipeline so new or updated policies discussed in meetings can be drafted and published with minimal manual effort

Why they built it

• Policies were scattered across Salesforce, Word docs, and PDFs — rarely updated, hard to find, and invisible to most employees
• The handbook was effectively "read once on day one, then forgotten" - no ongoing utility
• Notion was tried first but felt like "everyone else's Notion page" - couldn't match PayCaptain's branding or desired UX
• A real example of the problem: a referral policy with outdated payment terms led to an employee expecting payment they didn't receive
• Core value driver: radical transparency - PayCaptain wanted candidates, employees, customers, and partners to all access the same open information (e.g. maternity pay shouldn't be a secret)

Tools used

• Lovable - no-code/vibe-coding web app builder used to design and host the handbook
• Claude - AI underpinning the handbook's chat assistant
• n8n - automation layer connecting handbook updates and Slack workflows
• Slack - internal interface for the people bot; managers and employees query handbook content here
• Granola - AI meeting transcription tool; action items from calls trigger policy drafting automations

How you could build this yourself

1. Audit your current handbook - identify where policies live, when they were last updated, and whether employees actually access them
2. Sign up for Lovable - no coding required; start with a single landing page and your top-level navigation structure
3. Port your core policies - begin with high-traffic pages (parental leave, compensation, benefits) and build out from there; prioritise pages you'd want candidates to see before accepting an offer
4. Add branding and structure - use Lovable's customisation to match your company's visual identity; organise content by audience (new starters, managers, candidates, customers)
5. Embed an AI assistant - connect Claude (or another LLM) to your handbook content; critically, add a no-fabrication instruction - if the answer isn't in the handbook, the bot should say so and direct users to a named contact
6. Build the Slack bot (optional but high-value) - use n8n to connect Slack to your handbook; allow employees to query policies without leaving Slack
7. Connect your meeting tool (optional) - if you use Granola or a similar transcription tool, set up an n8n workflow so policy actions from meetings flow automatically into a Lovable draft for your review
8. Publish openly - link the handbook in job ads, interview invitations, and onboarding checklists; share with partners and customers to extend its value beyond the internal team
9. Set a publish habit, not a rewrite habit - the automation stack means your main job is reviewing AI-drafted updates and hitting publish, not manually rewriting docs

Source

Built by Emily Sivey, Head of People at PayCaptain. Presented as part of an Open Org "Shipped by HR" session. The live handbook is publicly accessible and linked via PayCaptain's careers page and job adverts.