Creating a Role That Never Existed in the World (opens in new tab)
Technical Writing at Toss has evolved from producing documents to designing organizational knowledge systems. The core argument is that code captures outcomes but not the context, decisions, and history behind them—information that both employees and AI need. Toss’s Technical Writing Chapter therefore aims to make knowledge discoverable, structured, and increasingly self-sustaining, with the long-term goal of making the chapter unnecessary. ### Why Code Alone Is Not a Sufficient Source of Truth - Code records what a system does, but rarely explains: - Why it was designed that way - Which alternatives were considered - What organizational history shaped the decision - Employees often reconstruct context from colleagues’ memories or years-old messenger threads. - AI faces the same problem: it understands general knowledge but lacks an organization’s internal context. - A true Single Source of Truth requires both the code and the surrounding knowledge that explains it. ### From Writing Documents to Bringing Knowledge to People - Early work focused on creating onboarding documentation for the frontend chapter. - The team recognized that well-written documents are ineffective if people do not actively read or find them. - Toss created “Parkssi,” a chatbot integrated into messaging tools and IDEs: - Users ask questions conversationally. - Answers are based on existing documents. - Sources are provided for verification. - Instead of requiring employees to search for documents, the knowledge reaches them where they work. - Teams that previously ignored documentation began requesting similar systems to reduce repetitive questions and capture implicit knowledge. ### From Documents to Knowledge Systems - A knowledge system collects scattered information from code, conversations, deployments, and other sources. - It structures that information so both humans and AI can understand and use it through questions and automation. - Properly designed systems: - Increase productivity across the organization - Reduce communication costs as the company grows - Help employees adapt quickly and consistently - Toss defines knowledge as verified information that helps people understand situations, make better decisions, and act within a specific context. ### The Four Responsibilities of Toss’s Technical Writing Chapter - **Build products:** The chapter develops and operates “Todok,” an internal knowledge-management platform. - **Lead organizational documentation:** TWs work directly with teams to collect and organize knowledge according to each group’s needs. - **Automate Technical Writing:** AI workflows and automation help employees create and review documentation with consistent quality. - **Shape documentation culture:** The chapter runs company-wide sessions, documentation guilds, and knowledge committees focused on making knowledge easier for AI and people to use. ### A Broader Definition of the Technical Writer - Although the job title remains “Technical Writer,” the role now resembles a knowledge-infrastructure designer. - The role expanded by following organizational problems rather than staying within a fixed professional boundary. - The chapter’s ultimate goal is for teams to create and manage knowledge independently, eliminating the need for a dedicated TW chapter. - The series will explore Toss’s knowledge product, failed documentation approaches, automation of TW work, and the path toward self-sufficient organizations. Technical writers can create greater value by designing systems that make knowledge continuously available—not merely by writing more documents. As AI reshapes every profession, expanding a role around the problems it must solve is presented as an opportunity for any discipline.