How it works

The Business Case
You are absolutely right—while standard automation could do this, an AI Agent is far superior here because of the "messy" nature of email.

Clients don't send jobs in perfect spreadsheets. They send emails saying, "Hi mate, need a Java Dev, £60k, London, usual spec attached."

A standard automation would break trying to read that. An AI Agent, however, acts like a human assistant:

It Reads: It reads the email body and any attachments, understanding context just like a consultant would.

It Structures: It extracts the salary, location, and skills, even if they are buried in a paragraph of text.

It Creates: It logs the job in Bullhorn as "Open" automatically.

It Writes: Crucially, it doesn't just copy the boring job spec to your website. It uses marketing intelligence to write a fresh, catchy advert before posting it.

Built Using

x
Outlook trigger
→ listMessagesMe
Monitor Client Inbox - The agent actively watches your Outlook inbox for emails from known clients containing keywords like 'New Role' or 'Job Spec'.
x
OpenAI action
→ openai_extract
Extract Job Data - The AI reads the email body and attachments, intelligently extracting the Salary, Location, Job Title, and Description into structured data.
x
Bullhorn action
→ createJobOrder
Create Job in Bullhorn - Creates a new 'Job Order' record in Bullhorn using the data extracted by the AI, setting the status to 'Open' immediately.
x
Quake action
→ quake_create_job
Sync to Quake Core - Backs up the new Job record into the central Quake database to ensure we have a master record of all live roles.
x
OpenAI action
→ openai_generate_text
Write Marketing Copy - The AI takes the technical job spec and rewrites it into an engaging, SEO-optimized job advert ready for the web.
x
SourceFlow action
→ post___sf_api_v1_jobs.json
Publish to Website - Posts the newly written advert directly to your SourceFlow website, making the job live for candidates to apply.