FDE Times: FDE vs implementation engineer — the line is the first line of new code
A 4 November FDE Times guide says implementation engineers follow a configure-and-hand-over playbook; FDEs are called when that playbook runs out and someone must write production-grade code the product does not yet have — after proving configuration cannot do the job, and naming who will run the code after you leave.
Two engineers can sit in the same client office putting the same product live, but only one is asked to write what the product cannot do yet. FDE Times’s 4 November guide contrasts Rocketlane-style implementation playbooks (configure, set up, hand over) with forward-deployed work that starts when no admin screen answers the customer’s next question.
Classify each requirement before opening an editor: warehouse roles and CSV import are usually configurable; a 15-minute sync from a legacy ERP that only exports fixed-width files is not. A single ask often mixes both — configure the minimum thresholds that exist, and write only the seasonal reorder logic the product lacks. Writing code for something the product already configures creates dual mechanisms that break on the next upgrade.
Before the first new line: prove configuration cannot do it, write down the configurable vs custom boundary for both sides, and name who will run the code after you leave. Prototype-fast is fine; production means idempotent upserts, logged parse failures, and an alert when the file stops arriving. Palantir’s Forward Deployed Software Engineer posting stresses custom applications and strong coding; Deployment Strategist verbs lean scope and present, not build.
Source: fdetimes.net