The paper introduces flash teams, a framework for rapidly forming, coordinating, and managing small, paid expert crowdsourced teams to accomplish complex, interdependent tasks such as user‑centered design, animation, and curriculum creation. Flash teams are expressed as sequences of modular, typed blocks (tasks) linked by input/output tags, each with a designated Directly Responsible Individual (DRI). The authors built Foundry, an authoring and runtime platform that lets end‑users compose blocks, recruit experts from oDesk, and automatically handle handoffs, notifications, elastic scaling, and pipelining of intermediate results. Computational techniques (STRIPS planning) enable on‑demand generation of novel team structures by searching a library of existing blocks. The paper presents four concrete flash‑team families (napkin‑sketch design, animation, education, and a composite online‑course platform) and reports a controlled field experiment comparing flash‑team workflows to self‑managed teams. Results show flash teams complete work in roughly half the active hours, with lower variance, less coordination overhead, and comparable quality. The discussion outlines limits (coordination challenges, need for trusted worker pools) and future directions (broader task domains, richer runtime support, automated hiring). The work positions flash teams as a scalable, computationally‑enhanced alternative to traditional expert crowdsourcing and to in‑person temporary organizations.
Topics
Flash Teams Concept: Modular, Computation‑Ready Expert Crowds
Definition of flash teams as linked sequences of modular tasks (blocks) that can be interpreted and manipulated by software, enabling dynamic assembly of paid expert crowds.
Blocks and DRIs: Structured Task Units with Explicit Responsibility
Explanation of the block abstraction (input tag, output tag, skill requirements) and the Directly Responsible Individual role that provides lightweight leadership within each block.
Case Study 1 – Napkin‑Sketch Design Flash Teams
Empirical example where a user‑centered design pipeline (low‑fi mockup → heuristic evaluation → high‑fi prototype → user testing) was executed by 3‑10 experts, achieving functional mobile apps within 18‑31 hours of work.
Case Study 2 – Animation Flash Team
Workflow that turns a script idea into a short animated video via scriptwriting, storyboarding, character design, and animation, illustrating flash teams in a creative, non‑engineering domain.
Case Study 4 – Composite Online‑Course Platform
Large‑scale composition of nine flash teams (design, education, animation) to build a multi‑course MOOC platform, demonstrating modular recombination and parallel execution.
Field Experiment: Flash Teams vs. Self‑Managed Teams
Controlled study with six napkin‑sketch teams (three flash, three control) showing flash teams required ~13 h of active work versus ~24 h for controls, with less variance, lower coordination load, and comparable quality.
Discussion of Limits and Future Work
Critical reflection on coordination challenges, need for trusted worker pools, potential for manager‑led variants, richer runtime correction, automated hiring, and broader task domains.