Project Management Portfolio

Producing my MFA thesis short, Nibble: an 18-week preproduction plan, a four-day shoot with one lead per day, a $4,750 Movie Magic budget with 33 of 44 lines covered without cash, and the changes and failures the plan absorbed. Accepted by three festivals; Best Student Film at Think Film.

The real documents behind the case study: the preproduction schedule, the stripboard with my board notes, a Day Out of Days, every budget line, and the risk and change log. Names and contact details removed.

From fourteen years of producing film and television: a call sheet is the day’s plan made public, and every mistake on it lands on the whole crew. A training sample for new ADs and producers checks tomorrow’s draft against the schedule, the cast list, the daylight and the 10-hour turnaround, and catches four errors before it goes out. Part of a call-sheet course built for every crew member.

An evaluation plan for a film set’s safety orientation, owned by the 1st AD and reported to the producer. Four levels, each with a measure, a timeline and a target; a ten-item set-walk checklist on days 1, 3 and 10; weekly reporting; and an honest account of the limits and the cost, using only what a small production already has.

Three film-crew training samples planned as one production, with one crew, brand and map style across all three. Each open question settled with a single ruling, every path tested before release, and a safety evaluation plan the 1st AD owns and reports to the producer.

Twelve sheets from one locked brief: the facts fixed and sourced first, four brand standards set down as tokens, an approval gate after each piece, and every problem found in render review logged with its fix.

Three courses for four audiences in two days. Process gaps became ten rulings, asked one question at a time. A $1,449-a-year licence was weighed against a free rebuild from the same storyboards, and every course launched with a live door to point at.

An eleven-page software course and a Rise 360 version, both delivered in one day. Every step was checked against the data before it was written, and differences from the plan were logged and resolved in favour of the app. When the course exposed thin sample data, the fix went into the product: a ruling, a 0.9.2 release, and the whole course re-shot from its scripts in one pass.

A training course built against a product still in testing, web and Rise versions in two days. All 18 findings from walking the lessons in the app were logged, and each real fault went up as a single decision: fix the product first, then teach it. Six fixes were released, tested and re-shot into the lessons before launch, and the final SCORM test caught a defect shared with an earlier project, fixed in both.

A feature asked for at 08:11 and released across the Character Creator, the Battle Tracker and two Owlbear extensions by 10:35. Four options were weighed, the result was held for review, the bug that prompted it was logged rather than chased, and a wrong answer about its limits was corrected in the open.

A rules change raised by playtesters and the team that reached the book, the Character Creator, the Battle Tracker and two Owlbear extensions in one morning, with every saved character migrated. It also covers a release held for a ruling when two parallel workstreams overlapped in one shared file, and what changed in the process because of it.

Three tools for the table built on one set of rules code, shipped as a release train across four products in two days: the Character Creator through six versions, the Battle Tracker through four, and two Owlbear Rodeo extensions synced alongside. A team request and a reported freeze were each turned around the same day, and every release was checked for drift across the apps.

Five weeks of changes to the Character Creator, each traced to where it came from: a playtester-driven rebuild of how creation buys gear, a language review, a rules-change pipeline that tells returning players what changed, and a wrong fix undone the same day.

Running a design, playtest, engineering, web development and language team as a roundtable with one decision-maker. A single source of truth, change control, versioned releases, and an AI assistant whose work gets checked like anyone else’s.

Mapping every dependency between the project’s tools from the source code, not from memory, and catching a live, published error along the way.

A playtester, a designer and an AI edit pass caught the same gap on their own. One person confirmed the match, the fix was simulated before anyone played it, and every ruling was made with the numbers in hand.

A ten-milestone rules rewrite run one decision at a time, with scope logged as it grew, every rule simulated, and the text and visuals shipped in separate reviewed commits.