Developer Unveils AI-Native Resume: Recruiter Bots Can Query Experience Instantly

A software engineer has created a groundbreaking open-source project that turns a traditional PDF resume into a live, queryable data source for artificial intelligence agents. The innovation allows hiring bots to bypass messy PDF parsing and instead request structured information on demand.

"I asked myself: why not make my resume AI-native? Not 'AI-friendly' marketing fluff, but something an AI agent can actually connect to and query in real time," said Leeson Wong, the project's creator. His solution, hosted on GitHub, implements an MCP server that AI agents can auto-discover and use.

The breakthrough comes as companies increasingly rely on AI to screen resumes. Traditional PDFs are notoriously hard for AI to parse due to layout noise and scattered information. Wong's MCP server eliminates these barriers by exposing a defined set of functions—such as search_resume, evaluate_fit, and get_career_summary—that return clean JSON results.

Background: The Rise of AI Screening and MCP

AI recruitment tools have exploded in usage. Human resources departments hand job descriptions to AI agents, which then attempt to match candidates. But these agents struggle with unstructured PDFs, often missing key context or extracting irrelevant data.

To solve this, Anthropic proposed the Model Context Protocol (MCP)—an open standard that lets AI agents connect to external data sources. MCP is self-describing: it provides tools with names, descriptions, and input schemas, allowing AI agents to understand how to use them without integration guides. Wong applied this protocol to his resume, effectively creating a real-time API for his professional history.

"The technical implementation is basic, but the product logic is what matters," Wong noted. "This isn't an AI-gimmick—it's a fundamental shift in how resumes are consumed."

What This Means: A New Recruitment Paradigm

The Resume Submission Flow Has Changed

The old process was passive: send a PDF, HR skims it manually, then maybe passes it to an AI for supplementary analysis. Wong's approach flips the script. Now, a recruiter gives candidates a link and an invite code. Their AI agent connects via MCP and decides what to query—and how to analyze the results.

"The initiative shifts from the human to the AI agent," explained Wong. "The agent can dynamically query based on what it cares about, rather than passively reading a fixed document." This allows for deeper, more relevant assessments. For example, an agent can instantly ask: "Evaluate fit for a senior full-stack role requiring React, Node.js, and big data experience" and receive a structured score with matched and missing skills.

Invite Codes = Reverse Filtering

Perhaps the most intriguing aspect is the invite code mechanism. Traditional screening is top-down: employers filter candidates. With this MCP server, candidates can control access—each query requires an invite code that the candidate issues. This means candidates can selectively grant or deny permission, forcing recruiters to prove their legitimacy before they can analyze the resume.

"The invite code turns the tables," Wong said. "It's a form of reverse filtering—the candidate decides who gets to see their structured data, not the other way around." This could reduce spam applications and improve the quality of recruiter-candidate interactions.

Key Capabilities in Action

After connecting to the MCP server, an AI agent performs structured queries that return clean JSON objects:

• Search for experience: search_resume({"query": "big data"}) returns relevant snippets with source modules and relevance scores.
• Job fit evaluation: evaluate_fit({"job_description": "..."}) returns a match score, a list of matched and missing skills, and a recommendation.
• Comprehensive assessment: get_career_summary({}) provides seniority level, total years, domains, core strengths, and top skills with proficiency ratings.

All responses are structured JSON—no PDF parsing, no web scraping. The AI agent processes data programmatically, eliminating errors caused by layout noise.

Implications for Job Seekers and Recruiters

This development could disrupt how resumes are written and shared. Instead of optimizing for keyword stuffing or ATS compatibility, candidates might soon create machine-readable endpoints that let AI agents explore their experience dynamically. Recruiters, in turn, could gain far richer insights than a static PDF can offer.

"We're moving from documents to APIs for hiring," Wong predicted. "Your resume becomes an endpoint that AI agents can call on demand." The open-source nature of the project invites community contributions and scrutiny, potentially setting a new standard for AI-native recruitment.

For now, Wong's project remains a proof of concept, but its implications are wide: if adopted broadly, it could reshape the entire recruitment pipeline—reducing friction for both candidates and employers while increasing the depth and fairness of candidate evaluation.