j4flmao R&D and Proof of Concept (PoC) Management
1. Skill Context
Focus: Managing the lifecycle of technological innovation, validating hypotheses rapidly, and successfully transitioning experimental code (PoC) into enterprise-grade production software. Triggers: r&d, poc, prototype, mvp, architecture, adr, innovation, enterprise-rd.
2. The "Valley of Death"
In enterprise environments, there is a notorious "Valley of Death" between the R&D Department and Global Operations.
- A Department builds a brilliant PoC in 3 weeks. It works perfectly in isolation.
- Global Operations rejects the PoC because it lacks monitoring, uses unapproved databases, fails compliance checks, and is undocumented.
- The PoC dies, and the Department's innovation effort is wasted.
This skill path provides the architectural and procedural patterns required to bridge that valley.
3. Core Principles of R&D
- Hypothesis-Driven Engineering: A PoC is not the first draft of an application. It is an experiment designed to answer a specific technical or business question (e.g., "Can Kafka handle 100k events/sec on our current hardware?").
- Documented Context: R&D decisions often look crazy to outsiders. Documenting why a decision was made (using ADRs) is more important than the code itself.
- Strict Timeboxing: A PoC must have a hard deadline. If it takes 6 months, it is not a PoC; it is a shadow IT project.
4. References
references/architecture-decision-records.md— How to write and manage ADRs.references/build-vs-buy.md— Framework for Technology Selection.references/poc-success-criteria.md— Defining NFRs and exit criteria.references/throwaway-vs-evolutionary.md— Architecture strategies for prototypes. (Batch 2)references/sandbox-environments.md— Cloud isolation and safety. (Batch 2)references/api-contract-mocking.md— Unblocking development. (Batch 2)references/poc-to-mvp-handoff.md— The Global transition process. (Batch 3)references/retrofitting-security.md— Paying back technical debt. (Batch 3)