軟體工程師 (前端) ｜ Software Engineer (Frontend)

【職位概述】 Tricuss (十論科技) 致力於顛覆傳統研發流程（Revolutionize RD）。旗艦產品「企業級 Co-researcher AI Agent」為自主虛擬科學家，融合 Agentic AI、大數據與物理模擬，能自主設計假說、推演，並自動檢索論文與內部資料以加速研發探索。 我們將系統與「物理基礎數位孿生（Physics-based Digital Twins）」深度結合，專為地端（On-premise）與封閉網路（Air-gapped）高資安客戶打造，實現最高千倍加速並減少 70% 實體實驗。目前我們與 NVIDIA、Qualcomm、Intel 及世界頂尖半導體與 AI 伺服器供應鏈的客戶們密切合作，邀您共同定義下一代研發標準。 身為前端工程師，您將是把強大但複雜的 AI 研發能力，轉化為使用者真正信賴的產品介面的關鍵推手。我們的使用者是嚴謹、精準的半導體與材料科學家——他們要的不是花俏的介面，而是能可靠呈現海量製程資料、複雜模擬結果與研究洞察，讓他們快速做出決策的工具。隨著 Co-researcher 平台功能擴張，您將負責用 React + TypeScript 開發並擴展多個資料密集的產品介面（內部管理後台、客戶端 Web App、代理工作流介面、報告生成），把工程級的資料密度與一流的使用體驗同時做到位，並與設計團隊共同打造可長期維護的 design system。 【主要職責】 產品介面開發與擴展： 使用 TypeScript + React.js + Next.js（含 App Router、SSR / SSG）開發與擴展多個產品介面，涵蓋內部管理後台、客戶端 Web App、AI 代理工作流介面與報告生成模組，確保高效能與一致的使用體驗。 資料密集介面打造： 開發資料密集的管理後台——大型表格（可達數千列）、複雜的過濾與排序、即時資料更新與串流，將複雜的模擬結果與研究數據以清晰、可操作的方式呈現給工程師使用者。 設計稿高還原與元件化： 與設計團隊定期走查設計稿，把 Figma 設計準確還原為高品質、可維護、可重用的 React 元件，雙向協作完善 design system 與元件規範。 前端效能優化： 針對大量 DOM 與即時資料場景，運用虛擬化（virtualization）、memoization、bundle size 分析與 React Profiler 等手段，持續優化渲染效能並排除潛在瓶頸。 前後端協作與整合： 與後端工程團隊（TypeScript + Node.js + MongoDB）緊密協作，串接 RESTful API，共同定義資料介面契約，確保功能完整且穩定的端到端體驗。 工程品質與可維護性： 撰寫清晰、可測試、可維護的程式碼，建立與維護單元測試（Vitest / Jest / Testing Library），並協助把前端某一產品領域的標準與最佳實踐沉澱下來。 === 【Position Overview】 Tricuss is on a mission to revolutionize RD. Our flagship enterprise-grade Co-researcher AI Agent is an autonomous virtual scientist that fuses Agentic AI, big data, and physics simulation to formulate and reason through hypotheses on its own, while auto-retrieving global academic papers and internal data to accelerate RD. Built on Physics-based Digital Twins for on-premise and air-gapped, high-security customers, it delivers up to 1000x acceleration and cuts physical experiments by 70%. We work closely with customers across NVIDIA, Qualcomm, Intel, and the world's leading semiconductor and AI server supply chains—come help define the next generation of RD standards. As a Frontend Engineer, you'll turn powerful but complex AI RD capabilities into product interfaces users genuinely trust. Our users are rigorous semiconductor and materials scientists who don't want flashy UI—they want tools that reliably present massive process datasets, complex simulation results, and research insights for fast decisions. As the Co-researcher platform expands, you'll build and scale multiple data-intensive interfaces with React + TypeScript (internal admin consoles, client-facing Web Apps, agent workflow UIs, report generation), pairing engineering-grade data density with a best-in-class experience and partnering with design on a maintainable, long-lived design system. 【Key Responsibilities】 Product interface development and scaling: Build and scale multiple interfaces in TypeScript + React.js + Next.js (App Router, SSR / SSG)—internal admin consoles, client-facing Web Apps, AI agent workflow UIs, and report generation—with high performance and a consistent experience. Building data-intensive interfaces: Develop data-intensive admin consoles—large tables (up to thousands of rows), complex filtering and sorting, real-time updates and streaming—presenting complex simulation results and research data to engineers clearly and actionably. High-fidelity design implementation and componentization: Review designs regularly with the design team, faithfully translate Figma into high-quality, maintainable, reusable React components, and collaborate to refine the design system and component standards. Frontend performance optimization: For large-DOM and real-time scenarios, optimize rendering and remove bottlenecks via virtualization, memoization, bundle size analysis, and the React Profiler. Frontend–backend collaboration and integration: Work closely with the backend team (TypeScript + Node.js + MongoDB), integrate RESTful APIs, jointly define data interface contracts, and ensure a stable end-to-end experience. Engineering quality and maintainability: Write clear, testable, maintainable code, build and maintain unit tests (Vitest / Jest / Testing Library), and help codify standards and best practices for a frontend product domain.