Press

Automated Resume-to-Posting Alignment Tool

Press your resume against any job posting and see what sticks.

Overview

Press is the automated version of the Tailoring Blueprint service. It takes a resume and a job posting as inputs, scores the match, identifies gaps, and generates tailoring recommendations with example bullet rewrites. Dual-purpose: O.J. uses it to speed up service delivery, and clients can use it as a self-service tool.

Named after a citrus press — you press the resume against the job posting and extract the juice (the match score, the gaps, the actionable insights).

Problem

Clients who complete a Fresh Squeeze resume review often ask "I put my resume out there but I'm not getting any bites. What do I do?" The answer is usually: tailor the resume to each job posting. But the manual Tailoring Blueprint process takes 15-20 minutes per posting, and clients don't always know how to do it themselves even with the DIY one-pager. Press automates the repeatable parts and makes the judgment-heavy parts faster.

Architecture

Input Layer

Two inputs, both plain text for the MVP. User pastes their resume content and pastes the job posting. File upload (docx, PDF) is a v2 feature. The tool should handle both clean resumes and messy LinkedIn copy-paste, so it includes light normalization on input.

Processing Layer (Three Stages)

Stage 1: Job Posting Parser (rule-based)

Extracts structured data from the posting: required skills and tools (hard skills), soft skills and qualifications, experience level expectations, responsibilities, and domain context. Mostly rule-based with some NLP. Tech job postings follow predictable patterns ("X+ years of experience with Y", "proficiency in Z", "familiarity with W"). A curated taxonomy of common tech skills and tools gets smarter over time as more postings are processed.

Stage 2: Resume Analyzer (rule-based)

Parses the resume into sections and extracts the same categories: skills mentioned, tools and technologies, quantified achievements, experience domains. Maps these against what Stage 1 found. Produces the match score as a breakdown: hard skills coverage, experience alignment, keyword presence, and an overall composite.

Stage 3: Tailoring Engine (LLM-powered)

Takes the gaps identified in Stage 2 and generates actionable output: bullet rewrites that naturally incorporate missing keywords, suggestions for which existing bullets to modify vs. where to add new ones, flags for genuine gaps (things they can't honestly claim) with advice on how to address them in a cover letter or interview. The Tailoring Blueprint methodology is encoded as the system prompt and guardrails. Includes an authenticity check to make sure the tailored resume still sounds like the person.

Output Layer

The report: match score with breakdown, prioritized list of changes (highest impact first), side-by-side bullet suggestions (original and tailored version), gap analysis with honest "you don't have this, here's how to handle it" notes, and a short checklist.

Match Score Design

The score is a composite with a category breakdown:

Category	Description	Weight
Hard skills coverage	Percentage of required tools/technologies present	High
Experience alignment	Do years and level match the posting?	Medium
Keyword presence	Are the specific terms from the posting in the resume?	Medium
Domain fit	Does the industry/domain experience match?	Low

A single headline number (e.g., "62% match") with the breakdown visible underneath. The headline number gets attention, the breakdown drives action.

Important nuance: a resume could match 90% of keywords but miss the one dealbreaker requirement, or match 40% of keywords but be a strong candidate because their experience is highly relevant even if the terminology doesn't line up. The score should flag must-have misses prominently regardless of the overall percentage.

Hybrid Engine and Cost Structure

The hybrid approach (rule-based scoring + LLM suggestions) creates a natural freemium model:

Free tier: Match score and gap identification. Rule-based, no API costs. Teaches the user something just by showing them which keywords they're missing and where.
Premium tier: LLM-powered bullet rewrites, reframing suggestions, and cover letter guidance. This is the service O.J. delivers in paid engagements or gates behind a paid self-service tier.

This aligns with the anti-gatekeeping philosophy. The score itself is educational and free. The premium layer is where the judgment and craft live.

Format

Web app hosted on learnwithoj.com. Clients are software and infrastructure engineers (not command-line-averse but a web interface removes friction). Aligns with the long-term "Khan Academy for adults in tech" vision where free tools drive traffic to paid services.

Development path: prototype as an artifact first to nail UX and logic, then pull into the site.

Integration Points

Press ← Tailoring Blueprint Methodology

The Tailoring Blueprint process doc (in processes/) defines the steps Press automates. The system prompt for Stage 3 should encode the methodology's guardrails: authenticity checks, reframable vs. true gap distinction, and the "teach a man to fish" philosophy.

Press ← Fresh Squeeze

Press is designed to be used after a Fresh Squeeze review. The resume should be "ready for prime time" before tailoring begins. Press could eventually accept a Fresh Squeeze report as additional context to avoid re-analyzing the resume's structural issues.

Press → Zest

Press engagement data (which job titles and skills are most commonly searched, which gaps appear most often) could feed Zest's seed engine for content ideas. "60% of Press users are missing Kubernetes experience from their resumes" is a Hands-On Tuesday post waiting to happen.

Implementation Phases

Phase 1: Static Prototype

Build an interactive artifact that demonstrates the input/output flow. Hardcoded example data. Validates the UX and report format. No backend.

Phase 2: Rule-Based Scoring

Build the job posting parser and resume analyzer. Produce the match score and gap identification. No LLM calls yet. This is the free tier.

Phase 3: LLM-Powered Suggestions

Add the tailoring engine using the Claude API. Bullet rewrites, reframing suggestions, cover letter guidance. This is the premium tier.

Phase 4: Web App Deployment

Deploy to learnwithoj.com. Authentication if gating the premium tier. Usage tracking.

Phase 5: Tech Taxonomy

Build a curated, growing taxonomy of tech skills and tools that improves keyword extraction accuracy over time. Could be community-contributed eventually.

Open Questions

How should the match score handle dealbreaker requirements that are missing? A prominent "critical gap" flag? Or does the score drop disproportionately?
Should Press store previous analyses so clients can track improvement across tailoring iterations?
What is the right boundary between what Press does automatically and what still requires O.J.'s judgment? The authenticity check is hard to automate.
Should Press support comparing a single resume against multiple postings at once? Clients often apply to several similar roles.
How does Press interact with hiring-scout (the job search automation tool)? If hiring-scout finds postings, Press could auto-score them against the client's resume.
Is there value in a "reverse Press" mode where a client pastes a job posting and Press suggests what resume bullet topics they should develop before applying? More of a career planning tool.
Should the DIY one-pager from the Tailoring Blueprint methodology be built into Press as an interactive tutorial?

Press | Learn with O.J. | learnwithoj.com

Press Design Doc