How to Automate Employee Onboarding with Paradime, OpenClaw, and Google Workspace

Every data team has been there. A new hire joins on Monday, and suddenly you're scrambling — digging through Confluence pages last updated in 2022, pinging three different Slack channels to find out who owns the warehouse credentials, and hoping that the "onboarding checklist" somebody bookmarked still exists. The result? A confused new team member, a frustrated manager, and a first week that feels more like an archaeological dig than a productive ramp-up.

In this guide, you'll build an automated employee onboarding pipeline using Paradime, OpenClaw, and the Google Workspace APIs. By the end, you'll have a system that reads new hire details from a Google Sheet, generates a role-specific onboarding checklist, sends a personalized welcome email, and schedules orientation meetings — all triggered automatically when a new hire record appears.

The Onboarding Problem: Stale Docs, Missing Context, Tribal Knowledge

Before we jump into the solution, let's be honest about the problem.

According to HiBob research, 64% of employees are likely to leave a new job within their first year after a negative onboarding experience. And HR Dive reports that 55% of organizations consider their ability to use preferred tools and documents in onboarding as fundamentally broken.

Here's what that looks like on the ground:

Figure 1: Manual vs. automated onboarding — the path from chaos to confidence.

The goal isn't to replace the human touch of onboarding. It's to automate the mechanical parts — the checklist generation, the email drafting, the calendar scheduling — so that humans can focus on the mentoring, the culture building, and the actual knowledge transfer.

What Is Paradime?

Paradime is an AI-native data platform that replaces dbt Cloud™. It brings the "Cursor for Data" experience to analytics and data engineering teams, combining an AI-powered IDE, production pipeline orchestration, and warehouse cost optimization into a single platform.

For this project, we care specifically about Bolt — Paradime's scheduler and orchestration engine. Bolt lets you:

• Run dbt™ pipelines on schedules, on Git merges, or via API triggers

• Monitor run history with Summary, Console, and Debug logs

• Set up notifications via Email, Slack, or Microsoft Teams

• Achieve 99.9% uptime with built-in CI/CD

Bolt's API trigger capability is what makes this automation possible. Instead of running on a cron schedule, our onboarding pipeline fires when an external system (in our case, the OpenClaw agent) calls the Paradime GraphQL API.

What Is OpenClaw?

OpenClaw is an open-source AI agent that runs locally on your hardware and orchestrates tasks across chat apps, files, the web, and your operating system. It isn't an LLM itself — instead, it connects to models like Claude or GPT via API and uses skills to act.

Key capabilities for our use case:

• Skills & Plugins: Extend with community skills or build your own custom skills

• Google Workspace integration: Read/write Google Sheets, send Gmail, manage Google Calendar events

• Cron jobs and proactive automation: Schedule background tasks that run autonomously

• Persistent memory: Remembers context across sessions, making it smarter over time

• Full system access: Read files, run shell commands, execute scripts

OpenClaw acts as the orchestration brain of our onboarding workflow — it reads the new hire data, generates the checklist, sends the email, and schedules the meetings.

Architecture Overview

Before diving into setup, here's how all the pieces fit together:

Figure 2: End-to-end onboarding automation flow — from HR data entry to scheduled orientation.

Setup: openclaw + Google Sheets API + Gmail API + Google Calendar API

Prerequisites

Before you begin, make sure you have:

• Node.js 22+ installed

• A Google Cloud Platform project with billing enabled

• A Paradime account with Bolt plan (pricing)

• An LLM API key (Anthropic, OpenAI, or OpenRouter)

Step 1: Install OpenClaw

Run the onboarding wizard to configure your LLM provider:

This will walk you through setting your default model, API keys, and gateway configuration.

Step 2: Enable Google APIs

In your Google Cloud Console:

1. Navigate to APIs & Services → Library

2. Enable the following APIs:

3. Go to APIs & Services → Credentials

4. Create an OAuth 2.0 Client ID (Desktop Application type)

5. Download the credentials JSON file

Step 3: Configure Google Workspace Skills in OpenClaw

Install the Google Workspace skill (commonly called gog / gogcli):

Create the custom onboarding skill directory:

Step 4: Set Environment Variables

Create or update your ~/.openclaw/.env file:

You can also configure these in your ~/.openclaw/openclaw.json:

Security tip: Never commit .env files or credentials to Git. Use a secrets manager like 1Password or store them as environment variables on your host machine. OpenClaw supports secret refs for production setups.

Step 5: Generate Paradime API Keys

1. Log in to Paradime

2. Navigate to Account Settings → Workspace Settings

3. Scroll to the bottom and click Generate API Keys

4. Name the key (e.g., onboarding-automation)

5. Grant the Bolt Schedules Admin capability

6. Store the API Key and API Secret securely — you won't be able to retrieve them again

Script: Read New Hire Details, Generate Checklist, Send Welcome Email, Schedule Orientation

The OpenClaw Skill Definition

Create the skill file at ~/.openclaw/workspace/skills/employee-onboarding/SKILL.md:

The Automation Script

For more programmatic control, create a helper script at ~/.openclaw/workspace/skills/employee-onboarding/onboard.sh:

Make it executable:

Testing the Skill Locally

You can test the complete workflow by messaging your OpenClaw agent:

Or test individual components:

Bolt Schedule: API Trigger on New Hire

Now let's connect the automation to Paradime Bolt so it can be triggered programmatically whenever your HR system detects a new hire.

Creating the Bolt Schedule

1. Navigate to Bolt in the Paradime UI

2. Click Create New Schedule

3. Configure:

4. In Trigger Types, select Scheduled Run and click the cron toggle to OFF (API-only trigger)

5. In Command Settings, add your dbt™ command:

6. In Notification Settings, add:

7. Click Deploy

Triggering via the Paradime GraphQL API

When the OpenClaw agent detects a new hire, it can trigger the Bolt schedule programmatically:

A successful response looks like:

You can also pass custom commands or a specific branch at runtime:

Connecting Bolt to OpenClaw

Add the API trigger to your onboarding skill by including a curl call in the script or by instructing the agent to call the Paradime API after processing each new hire. This ensures that any dbt™ models tagged with onboarding (e.g., models that update an employee_directory table or refresh onboarding metrics) run immediately after a new hire is processed.

Figure 3: Complete Bolt + OpenClaw integration — the new hire triggers both the agent workflow and the dbt™ pipeline.

Environment Variables Reference

Here's a consolidated reference of all environment variables used in this project:

Storage Locations (by precedence)

OpenClaw resolves environment variables in this order:

1. Process environment (parent shell/daemon) — highest precedence

2. .env in current working directory — does not override existing vars

3. Global .env at ~/.openclaw/.env — does not override existing vars

4. Config env block in ~/.openclaw/openclaw.json — applied only if missing

5. Login-shell import (optional, via OPENCLAW_LOAD_SHELL_ENV=1) — lowest precedence

For production deployments, use OpenClaw Secret Refs or a tool like 1Password Service Accounts to avoid storing plaintext credentials.

Monitoring and Debugging

Once your automation is running, you'll want visibility into both sides: the OpenClaw agent and the Paradime Bolt pipeline.

Monitoring OpenClaw

Run the diagnostic command to check the health of your OpenClaw setup:

Healthy signals to look for:

• Runtime: running

• RPC probe: ok

• No blocking config/service issues

• Connected/ready channels

Monitoring Paradime Bolt

Paradime provides three levels of run logs for every Bolt execution:

To access run logs:

1. Navigate to Bolt in the Paradime UI

2. Click on the new-hire-onboarding-trigger schedule

3. Select a specific run from the Run History panel

4. Scroll to Logs and Artifacts

5. Click on the specific command (e.g., dbt run) to view Summary, Console, or Debug logs

Setting Up Alerts

Configure Bolt notification settings to get proactive alerts:

• Email: Workspace-wide notifications for all team members

• Slack: Per-schedule notifications with customizable Alert Templates

• SLA Thresholds: Get notified if a run exceeds expected duration

Bolt System Alerts (workspace-level) also cover:

• Parse errors

• Out-of-memory runs

• Git clone failures

• 24-hour run timeouts

Figure 4: Bolt monitoring and alerting flow — from run status to root-cause analysis.

Troubleshooting Common Issues

OpenClaw Issues

The golden rule of debugging OpenClaw: Check gateway.err.log first. Not your config, not your code — the error log. You can find it at ~/.openclaw/gateway.err.log.

Paradime Bolt Issues

Google API Issues

dbt™ Models for Onboarding Metrics

To track the effectiveness of your onboarding automation, you can create dbt™ models that aggregate onboarding data. Here's a taste of what the implementation looks like:

And a metrics model to track onboarding velocity:

These models are tagged with onboarding, so they'll automatically run when Bolt triggers dbt run --select tag:onboarding after each new hire is processed.

Bringing It All Together: The Complete Workflow

Here's the full end-to-end flow from HR action to a fully onboarded new hire:

Figure 5: The complete automated onboarding pipeline — from HR entry to data warehouse refresh.

Coverage Checklist

With this setup, you achieve near-100% coverage across the onboarding workflow:

Wrapping Up

The gap between "we should automate onboarding" and actually doing it usually comes down to integration complexity. By combining Paradime Bolt for pipeline orchestration, OpenClaw for intelligent task execution, and the Google Workspace APIs for the actual business actions, you get a system that:

1. Eliminates stale docs — the checklist is generated dynamically based on the role, not copied from a two-year-old Google Doc

2. Removes tribal knowledge dependency — the process is codified in the OpenClaw skill, not locked in someone's head

3. Provides full context — new hires get a personalized email, a tailored checklist, and pre-scheduled meetings from Day 0

4. Scales effortlessly — whether you're onboarding 1 person or 50, the same automation handles it

5. Creates accountability — every run is logged in Bolt with Summary, Console, and Debug logs, plus Slack/Email alerts on failures

The best part? This entire setup is version-controlled, auditable, and self-documenting. The OpenClaw skill definition is the documentation. The dbt™ models are the metrics. The Bolt schedule is the process.

No more stale docs. No more tribal knowledge. No more "ask Sarah."

Resources

• Paradime Documentation

• Paradime Bolt – Creating Schedules

• Paradime Bolt API

• Paradime API Key Generation

• OpenClaw Documentation

• OpenClaw Skills Guide

• OpenClaw Custom Skills

• OpenClaw Troubleshooting

• OpenClaw Gmail Integration