PROJECTS / INDEPENDENT WORK

Career Fit

Turning career evidence into a decision-support system

An independent application that turns career history, skills, project evidence, and job descriptions into structured context for fit analysis, resume strategy, gap identification, and more deliberate application decisions.

← All projects

The problem

Job searching often depends on fragmented information: a resume, LinkedIn profile, remembered accomplishments, job descriptions, and ad hoc notes about strengths or gaps. That makes it difficult to evaluate fit consistently, tailor applications efficiently, or maintain a coherent professional narrative across many opportunities.

My role

Conceived and built as an independent end-to-end product, including product strategy, data model, user workflows, application architecture, AI integration, evidence structure, resume-preparation logic, interface design, and iterative development.

The approach

Technology & capabilities

Skills demonstrated

Outcomes

The problem was not writing resumes

Career Fit began with a practical frustration: tailoring a strong application for each job required far more than editing a few resume bullets.

A good decision depended on understanding:

Traditional resumes were too compressed to hold all of that context, and starting from memory for every application was inefficient and inconsistent.

I wanted a system that could preserve the underlying evidence first, then use that evidence to support better application decisions.

Building a professional narrative as structured context

One of the first ideas behind Career Fit was the Candidate Narrative.

The goal was to capture the professional identity that sits behind the resume: the themes, working style, strengths, positioning, leadership approach, technical breadth, and context that help explain why a particular experience matters.

That narrative became a reusable source of truth for job analysis.

Instead of asking an AI model to infer a career from a resume alone, Career Fit could supply richer context deliberately.

Evidence, not just claims

The application also separates narrative from evidence.

Career accomplishments, projects, skills, examples, and responsibilities can be stored as structured evidence records and then surfaced against specific job requirements.

This makes it possible to distinguish between:

“I think I can do this.”

and:

“Here is the evidence that demonstrates it.”

That distinction became especially valuable when a job description appeared to expose a gap that was actually just missing documentation.

From job description to fit analysis

Career Fit analyzes a job description against the candidate narrative, profile, and available evidence.

The result is not simply a single match score.

The application can identify:

That makes the analysis more useful than a generic percentage because it helps explain why a role is or is not a good fit.

Separating analysis from preparation

A key design choice was to separate understanding the opportunity from preparing the application.

The analysis phase asks:

“What does this job need, and how well does the evidence support it?”

The preparation phase asks:

“Given that analysis, how should the resume and application strategy change?”

That separation allows Career Fit to preserve a stable professional identity while still adapting the presentation to each opportunity.

AI as a reasoning layer, not the data store

AI plays an important role in Career Fit, but it is not the application’s memory.

The underlying narrative, profile, job data, evidence, and strategy structures are stored explicitly in the application.

AI is then used to interpret, compare, synthesize, and suggest.

This makes the workflow more repeatable and gives the application a structured foundation that does not depend on a model remembering previous conversations.

Building a bridge to external AI work

Career development does not happen entirely inside one application.

Useful insights can emerge from conversations with external AI tools, resume reviews, interview preparation, and other analysis.

Career Fit therefore includes structured import/export concepts such as context packs and AI update packets.

These workflows allow an external conversation to propose new evidence or narrative updates in a controlled format that the application can validate and merge.

The goal is to make AI collaboration portable without surrendering the application’s structured source of truth.

Human judgment stays central

Career Fit does not treat model output as authoritative.

AI-generated analysis is useful because it can surface patterns quickly, but the final interpretation still depends on judgment.

A requirement may look like a gap because the evidence has not been documented yet. A technical keyword may matter less than the underlying capability. A job may appear numerically strong while being a poor fit for the kind of work someone actually wants to do.

The application is designed to support those decisions, not make them automatically.

Building the product with AI

Career Fit has also become a practical demonstration of AI-enabled software development.

The product has been developed iteratively with coding agents assisting with implementation, refactoring, interface work, and feature development.

The product direction, architecture, business logic, evidence model, workflows, and validation remain driven by my own decisions.

That combination has allowed the application to evolve quickly while still reflecting a clear product point of view.

An application built around the real decision

Career Fit is not ultimately about generating a resume.

It is about answering a more useful question:

Given who I am, what I have actually done, and what this role requires, how should I approach this opportunity?

Everything else in the application follows from that.