Overview & Challenge: MiruList centers the product on group decision-making rather than personal cataloguing. Collaborative watchlists let members add films, vote, and converge on a choice. The platform runs the full movie‑night flow: schedule a night, collect nominations, manage RSVPs, rank candidates based on who’s actually coming, and flow outcomes into watch history with a post‑night rating prompt. Guests can participate without accounts via signed owner/member links, making it easy to include friends with zero friction.

Approach & Implementation: MiruList uses a modern TypeScript stack and focuses on the experience rather than incidental mechanics. Streaming availability is aggregated from a user’s Plex library and subscriptions and computed ahead of time so filters and cards are fast and never blocked by live calls. Two independent rankers support discovery and decision‑making: “what should we add?” and “what should we watch tonight?”, factoring in group preferences, availability, and attendance. Imports and subscriptions (RSS, TMDB lists, CSV, Plex history) keep lists fresh, while notifications and an activity feed keep everyone aligned without noise.

Role & Responsibilities: As a Product Lead and Software Developer, I own the product direction and delivery. I designed a differentiating AI agent development pipeline with four codified roles (Refiner → Planner → Implementer → Reviewer), a kanban that executes for automated handoffs, and human‑in‑the‑loop checkpoints that hold previews for deliberate decisions. I established a documentation lifecycle (Living/Accepted/Proposed/Reference) and a decision log that records what we declined and why. The result is a product that scales collaboration, respects privacy, and makes movie nights effortless.

TypeScript

Portfolio

MiruList.app

My Coaching Toolbox