Case Study: Automating Job Offer Screening for a Consulting Firm
The Challenge
Every day, around 15 project proposals land in this firm's inbox. Not one of them is low effort. Each one needs to be read, evaluated, priced, and matched to the right person.
That is not just admin work. That is decision-making.
And they were spending over 2 hours a day doing it manually, before even deciding whether the opportunity was worth pursuing.
I talk to a lot of teams in similar situations. The pattern is always the same. Leads and opportunities are coming in, but what happens next is slow, inconsistent, and hard to track.
The real issue here was not volume. It was that high-value opportunities were being treated exactly the same as low-quality ones.
Everything went through the same inbox, the same process, the same manual effort. That is where things start to break. Opportunities were delayed. Some were missed entirely. The team had no consistent scoring framework, and the manual handoff into their Monday.com board added another step where things could slip.
The Approach
The goal was not to process emails faster. It was to eliminate the need for manual triage entirely.
This is a pattern I see often. Teams try to automate small tasks while the actual bottleneck stays untouched. If you automate the wrong thing, you just move chaos faster (I wrote more about that in "Should You Automate This?").
To solve this properly, the system had to do more than move data. It had to make decisions.
That meant:
- •Parsing unstructured proposal emails and extracting structured data reliably
- •Applying a pricing filter before spending any AI budget on deeper analysis
- •Scoring each opportunity across dimensions the firm actually cares about
- •Matching the right internal consultant to the right type of work
- •Creating the item in Monday.com automatically, fully populated, with the deadline already set
The system had to be opinionated. Not just "here is the data," but "here is our recommendation and why."
The Solution
We built a multi-stage AI pipeline triggered by incoming emails via Azure Logic Apps, running on n8n.
In simple terms, every incoming proposal goes through a decision pipeline before a human ever sees it. This is exactly where most pipelines break down in practice (I explain this in more detail in "Why Your Leads Stop Responding").
When a new proposal arrives, the workflow fetches the full job posting, removes the HTML noise, and extracts structured fields such as role title, location, duration, hourly rate, deadlines, and requirements.
A first AI check looks at compensation immediately. If the net rate after fees falls below the firm's minimum of €25 per hour, the proposal is logged and closed. No further processing. No wasted tokens.
For proposals that pass this filter, a second layer evaluates content fit and role classification. It determines whether the work aligns with the firm's positioning and the type of consultant they would deploy.
A third scoring layer produces a full breakdown of the opportunity. This includes fit, commercial upside, win probability, required effort, risk, the underlying client need, and even a suggested alternative framing if relevant.
If the proposal is marked as a GO or a borderline case worth reviewing, the system queries the internal team database and identifies the best-suited consultant. It compares the project's core discipline against each team member's expertise, sector focus, and keyword profile.
Finally, the workflow creates a fully populated item in Monday.com. Title, assigned consultant, scoring breakdown, deadline, and summary are all filled in automatically. No manual input required.
The Results
The biggest shift was not speed. It was focus.
Only about 15 percent of proposals now ever reach a human, and those are the ones worth pursuing.
The system handles the full pipeline. Intake, screening, scoring, matching, and board management all run automatically.
- •Proposals are processed within minutes instead of hours
- •Every opportunity is evaluated using the same consistent framework
- •The Monday.com board is always up to date, with the right person already assigned
Impact: Around 2.5 hours of senior time saved every day. That equals roughly 50 hours per month. More than a full workweek reclaimed, every month.
But more importantly:
- •No high-quality opportunity is delayed
- •No proposal gets ignored
- •The team only spends time where there is real upside
Same volume of incoming work. Completely different outcome.
The difference is not the opportunities. It is the system around them.
Closing
Most teams do not realise how much time and opportunity is lost in these in-between processes until you map them out properly.
If you want, send me:
- •your website
- •how leads or proposals come in
- •what happens after
I will point out where things are slowing down and what I would automate first.