Next Mission Group Chrome Extension

I built an MVP Chrome extension for Next Mission Group focused on a practical translation problem: helping veterans connect military experience to civilian job opportunities in language that hiring systems and job seekers can actually use. The project sits at the intersection of product design, scoring logic, applied LLM use, and careful systems architecture.

Highlights

• Built a Chrome-extension MVP that translates military experience into civilian job fit.
• Combined ontology mapping, scoring logic, and LLM-assisted explanations in one workflow.
• Prioritized production constraints (auth, API security, provenance) over demo-only prototypes.

Problem

Military experience often does not map cleanly onto civilian job descriptions, especially on mainstream job boards. Veterans may have relevant skills and experience, but the language, credential expectations, and occupation structure on the civilian side can make that experience hard to interpret or surface.

The product goal was to make that translation more immediate and usable inside the job-search workflow itself.

Approach

The extension uses a Manifest V3 Chrome architecture with a side panel experience. Users can connect their account, maintain a profile, open a job posting on supported sites, and request an analysis of fit between the posting and their background.

Under the hood, the product combines:

• job-page detection and scraping for sites including LinkedIn, USAJOBS, Workday, Greenhouse, and Lever
• a user profile layer stored through the extension and synced to the server
• ontology-backed mapping between military roles, civilian occupations, and transferable skills
• heuristic scoring and explanation logic
• LLM-assisted narrative output where it adds value

Methods

The current architecture is notable for a few reasons:

• Production APIs run on Wix (Velo) rather than requiring a separate always-on Node service for the MVP path
• the extension uses a short-lived JWT connect flow tied to Wix Members authentication, which lets the extension act like a native client without exposing keys in the browser
• the backend keeps provider API keys server-side in Wix Secrets or local environment variables, never in the extension
• the ontology roadmap separates a canonical backbone from an inferred layer, so LLM-generated alignments can be reviewed and traced rather than silently replacing authoritative identifiers

I like this project because it is not just a model-in-a-notebook. It required decisions about product flow, authentication, operational tradeoffs, and how to use LLM output in a system that still needs clear provenance and guardrails.

Outcome / Impact

The project is still in progress, but the MVP already defines a coherent product path: install the extension, connect to the Next Mission Group site, sync a profile, analyze a job, save results, and capture feedback. The docs also lay out a realistic path for extending the ontology, reviewing inferred mappings, and supporting richer military-to-civilian translations over time.

Because the product is still evolving, I have intentionally kept this case study focused on architecture and product thinking rather than impact metrics. [NEEDS INPUT] if you want to add pilot status, usage, or outcomes from testing.

Tools

JavaScript, Chrome Extension APIs, Wix Velo, JWT-based auth, ontology design, heuristic scoring, LLM integration, product architecture documentation.