← Selected work

EpicRole

Two jobs can share a title and ask for very different things. I built EpicRole during my own job search to see where my experience lined up, decide which roles to pursue, and keep track of what happened next.

EpicRole desktop dashboardRole Radar and review queue on mobile
RoleProduct, Design & Build
StatusIn Development
Timeframe2026–Present
PlatformDesktop and Mobile
The Process

What makes a job worth opening?

What I check before I apply

Each time a job alert landed, I checked the same things. What would I own in this role? Did the salary make sense? Had I already passed on it? I used those questions to design the first review screen and tested it against postings from my own search.

Separate qualifications from preferences

I split suitability into five parts: role fit, work fit, company interest, freshness, and salary. A role at a company I follow might score well on salary but ask for experience I haven’t built yet. Keeping the reasons visible lets me make that call.

Decide where AI earns its cost

I use rules to compare straightforward facts like pay ranges and posting dates. I reserve AI for the harder comparison between a posting’s requirements and my experience. When the requirements are missing, the assessment can stay unknown.

Outcomes

One job search in practice.

I use EpicRole to manage my own search while I build it. These figures come from my workspace on November 4, 2026.

678Jobs added during the first 30 days of development.
301Jobs found through searches of my target companies during the same period.
9Epic Roles that most closely matched what I wanted in my next job.
01 / Discovery

My preferences shape the shortlist.

I can follow companies I care about without limiting discovery to those employers. In the job library, I put details like salary and status on the card so I can scan the list and record a decision in the same place.

02 / Suitability

A closer look at one match.

An Autodesk posting shows why the breakdown matters. Its salary range met my target. The role also asks for hands-on ML platform experience. Seeing that gap beside the reasons to apply is more useful to me than a single score.

03 / Iteration

What I’m watching next.

A posting can close while an application is still in progress. I keep source verification separate from my application history so checking a link can’t erase a decision. My next test is discovery quality: how often does a scan turn up a role I haven’t seen before that I’d actually apply for?