Project 3: AI Workflow
You are a team of interns at Red Circle Consulting, a company that helps clients automate their workflows and make data-driven decisions using AI.
Your client, Pathfind Careers, is a small career-coaching and recruiting startup that helps job seekers identify roles that fit their skills and experience. Pathfind wants to automate parts of their consulting process using AI, making it easier and faster to guide clients toward suitable job opportunities.
Your team will design and present a prototype tool that demonstrates how AI can support both the candidate and career advisor in the job search process.
Create an AI workflow in LogicStudio for a mutli-agent system with agents working in sequence that:
Analyzes a candidate’s CV, résumé, or personal website to extract skills, job titles, and experience.
Engages the candidate in a conversation to clarify missing / unclear information
Searches for relevant job opportunities on LinkedIn or Google Jobs, matching the candidate’s profile and preferences.
10/29 Update: Generated job links might not be real, because ChatGPT hallucinates, so:
If you can figure out how to ensure only real postings are listed, so much the better!
An alternate approach that you can use was demonstrated in class today, where you use the "File Output" component and create your own list of jobs that you use. This would not be acceptable in the real world, but is OK for our prototype.
Produces a summary output, showing:
A list of top-fit job openings
Key skills matched or missing
Create a Value-Added Feature: Use the generated list of viable jobs in some productive way that adds value for the client or customers/users
Your team should discuss this early on as a part of your team’s planning sessions; what is that you want to do, and how does it enhance the overall value of your product?
How you divide up the work that needs to be done between your agents is up to you and based on your problem decomposition. Part of the project involves determining what the agents can and should be doing.
Presentations will be during class on Wed 11/5/2025. Unless otherwise specified, all work is due by then as described below.
AI Workflow Prototype (LogicStudio.ai, downloaded .json file)
A working LogicStudio (or similar) project showing the automated process — from CV upload to job recommendations.
Use sample data: the best way to try it out is with your own CV/Resume, as you will be best equipped to see the quality of recommendations.
Submit:
Every member must create their own version of this (submitting your own .json file here by noon Fri 10/31), before the team converges on a shared final version.
Your final group version, along with your own version, must both be linked on your course website page for Project 3, along with a brief description and summary of differences.
Presentation (5 minutes, including a short demo)
You will pitch your prototype to a non-technical audience of clients from Pathfind Careers, who will evaluate your prototype for its usefulness, clarity, and potential for real-world application. Your presentation should include:
A brief overview of the client problem and your proposed solution
A live walkthrough of your AI workflow
A clear explanation of how automation saves time, improves efficiency, or enhances customer experience
Demonstrates added value
Submit your group's presentation slide deck by posting it on your course website page for Project 3.
STAR Story
Submit your STAR (Situation, Task, Action, Result) story by posting it on your course website page for Project 3, indicating how YOU contributed to your team’s efforts.
Keep the user journey clear — how does the job seeker move through your system?
Use mock CVs and sample job listings for testing.
Make your prototype interactive enough to demonstrate value, even if some components are simulated.
Think like a consultant — focus on automation impact and client value, not just tech.