Welcome to the exciting world of Parcha agents! In this guide, we’ll walk you through creating your first agent and watch it come to life. Buckle up, because you’re about to supercharge your KYB process!

The Anatomy of an Agent 🧠

Let’s start by looking at a sample agent configuration. This is where the magic begins:

agent_key = "public-bdd"
created_at = "2024-08-01"
agent_name = "Public Business Due Diligence"
agent_type = "kyb"

agent_description = "This agent is responsible for performing BDD tasks."

[[steps]]
step_name="Web Presence Check"
command_type="tool"
instructions="Verify if the business has a website."
command_id="kyb.web_presence_check"
data_loader_id="kyb.web_presence_data_loader"

[[steps]]
step_name="Adverse Media Check"
command_type="tool"
instructions='''Verify that the individual doesn't have any adverse media related to the following:
* Fraud
* Bribery
* Illegal Activities
* Terrorism
* Crime
* Counterfeit
* Corruption
* Political Involvement
* Financing
* Tax Evasion
* Mining
* Money Laundering
* Human Smuggling
* Human Trafficking
* Interpol
* OFAC
'''
command_id="kyb.screening.adverse_media_tool"
data_loader_id="kyb.adverse_media_webcheck"

[steps.data_loader_args]
prerequisite_loader_id = "kyb.web_presence_data_loader"

[[steps]]
step_name="High Risk Country Check"
command_type="tool"
instructions="Verify that the customer does not operate in prohibited / high risk countries."
command_id="kyb.high_risk_country_tool"
data_loader_id = "kyb.countries_data_loader"

[[steps]]
step_name="High Risk Industry Check"
command_type="tool"
instructions="Verify that the customer does not operate in prohibited / high risk industries."
command_id="kyb.high_risk_industry_tool_v2"
data_loader_id = "kyb.industries_data_loader"

This configuration defines the agent’s identity, purpose, and the steps it will take to perform its due diligence tasks. It’s like giving your agent a brain and a to-do list all at once!

The Journey of Your Agent 🌟

Now, let’s visualize what happens when you trigger the /startKYBAgentJob endpoint. Brace yourself for an epic adventure!

Here’s what’s happening behind the scenes:

  1. Validation Checkpoint: The schema and agents undergo a validation process. If any issues are detected, an error message is returned.

  2. Job Queuing: Upon successful validation, the job is queued in the asynchronous AI engine.

  3. Data Loaders Execution: The data loaders are executed to gather necessary information.

  4. Parallel Checks Execution: Checks are run in parallel to ensure efficiency.

  5. Final Assessment Generation: A final assessment is generated based on the results of the checks.

  6. Webhook Notification: If a webhook is provided, it is triggered with the job object to notify the user of the job completion.

Bringing Your Agent to Life 🎭

To set your agent in motion, you’ll use the /startKYBAgentJob endpoint. This is where the rubber meets the road, and your agent springs into action!

Stay tuned for our next section, where we’ll dive deep into the /startKYBAgentJob endpoint and all its powerful fields. You’ll learn how to communicate with your agent and set it on its mission to revolutionize your KYB process.

Remember, with Parcha agents, you’re not just doing due diligence – you’re embarking on an AI-powered adventure in compliance and risk management! 🚀🔍💼