{"id":810,"date":"2026-08-08T12:13:38","date_gmt":"2026-08-08T12:13:38","guid":{"rendered":"https:\/\/meta.ngochn.com\/?p=810"},"modified":"2026-08-15T14:28:51","modified_gmt":"2026-08-15T14:28:51","slug":"ensa-workspace","status":"publish","type":"post","link":"https:\/\/meta.ngochn.com\/index.php\/2026\/08\/08\/ensa-workspace\/","title":{"rendered":"Ensa Workspace"},"content":{"rendered":"
\n
\n
\n

Agenda<\/strong><\/p>\n

    \n
  1. Overview & Role:<\/strong> Project scope and my 100% solo ownership.<\/li>\n
  2. The Problem & Market Gap:<\/strong> Identifying the need for accessible financial reports.<\/li>\n
  3. Research & The Pivot:<\/strong> Shifting focus from retail to professional users.<\/li>\n
  4. UX Strategy & Key Decisions:<\/strong> Core principles for building trust in AI.<\/li>\n
  5. Final Solution & Designs:<\/strong> Interface screens and interactive flows.<\/li>\n
  6. Technical Execution & Roadmap:<\/strong> System architecture and rollout phases.<\/li>\n
  7. Business Impact:<\/strong> Driving MAU and subscription revenue.<\/li>\n
  8. Reflections & Takeaways:<\/strong> Lessons learned and navigating tight constraints.<\/li>\n<\/ol>\n<\/div>\n
    \n

    1. Project Overview<\/h4>\n

    An AI-powered research workspace for investors, where users interact with financial reports through natural language, receive AI-driven insights with verifiable citations, and publish living research reports.<\/p>\n<\/div>\n

    \n
    My Role & Contributions<\/h5>\n

    As a Senior UX Designer, I operated as the solo designer on this project, taking 100% end-to-end ownership. I acted as the de-facto Design Lead, driving the product vision, conducting research, and managing cross-functional alignment. This demonstrates my readiness and capability to lead complex, strategic AI initiatives.<\/p>\n<\/div>\n

    \n\n\n\n\n\n\n\n\n
    <\/th>\nResponsibility<\/th>\nDetail<\/th>\n<\/tr>\n
    1<\/td>\nEnd-to-end UX design<\/strong><\/td>\nFrom concept, landing experience, to complex workspace interactions.<\/td>\n<\/tr>\n
    2<\/td>\nInformation architecture<\/strong><\/td>\nSession management, project organization, document hierarchy.<\/td>\n<\/tr>\n
    3<\/td>\nAI interaction patterns<\/strong><\/td>\nChat UI, citation navigation, confidence indicators.<\/td>\n<\/tr>\n
    4<\/td>\nCross-functional leadership<\/strong><\/td>\nAligning BA, Engineering, and UXR teams to execute the product vision.<\/td>\n<\/tr>\n
    5<\/td>\nResearch Presentation<\/strong><\/td>\nPresented comprehensive research to the PO detailing market analysis, competitive landscape, and user journeys.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<\/div>\n

    <\/p>\n

    \n

    2. The Problem & Market Gap<\/h4>\n

    Financial reports are inaccessible to most investors.<\/h5>\n\n\n\n\n\n\n\n
    Pain Point<\/th>\nDetail<\/th>\n<\/tr>\n
    Barrier to Entry<\/strong><\/td>\nReports are filled with jargon, dry tables, no context.<\/td>\n<\/tr>\n
    Lack of Insight<\/strong><\/td>\nExisting tools extract data and draw charts, but don’t explain meaning.<\/td>\n<\/tr>\n
    No Interaction<\/strong><\/td>\nUsers can’t ask questions directly about a report.<\/td>\n<\/tr>\n
    Risk of Oversight<\/strong><\/td>\nEasy to miss red flags, anomalies, and hidden risks.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

    Competitive Positioning<\/strong>
    \nWe aimed to build the “NotebookLM for Financial Reports” specifically for the Vietnamese market.<\/p>\n\n\n\n\n\n\n
    Competitor<\/th>\nTheir Approach<\/th>\nOur Differentiation<\/th>\n<\/tr>\n
    FiinPro \/ WiChart<\/strong><\/td>\nData extraction and charts<\/td>\nAI explanation, interactive Q&A, natural language<\/td>\n<\/tr>\n
    Google NotebookLM<\/strong><\/td>\nGeneral-purpose document analysis<\/td>\nSpecialized for financial reports with domain expertise<\/td>\n<\/tr>\n
    AlphaSense<\/strong><\/td>\n$10K\u2013$40K\/user\/year<\/td>\nAffordable, Vietnam market-first, AI-native<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n
    \n

    Our moat:<\/strong> “Prepared context quality”. How well we pre-process and structure documents before AI ingests them is the real defensibility. Not the AI model itself.<\/p><\/blockquote>\n<\/div>\n<\/div>\n

    <\/p>\n

    \n

    3. Research & Target Users<\/h4>\n
    \n
    \n

    Phase 1: Retail Investors<\/strong><\/p>\n

      \n
    • Age 22\u201340, no deep finance background<\/li>\n
    • Overwhelmed by financial report jargon<\/li>\n
    • Want “easy answers” from complex data<\/li>\n<\/ul>\n

      Key research insight:<\/strong> Retail users want “cheap & easy” (commoditized by ChatGPT)<\/p><\/blockquote>\n<\/div>\n

      \n

      Phase 2 (The Pivot): Professionals<\/strong><\/p>\n

        \n
      • Research analysts, advisory teams, brokers<\/li>\n
      • Content publishers who write & distribute research<\/li>\n
      • Clear pain points, willingness to pay<\/li>\n<\/ul>\n<\/div>\n<\/div>\n
        \n

        Key research insight:<\/strong> Professional users have distinct workflow pain points and measurable willingness to pay for specialized tools. This forced a full redesign from a simple chat to a multi-session, project-based environment.<\/p><\/blockquote>\n

        \"\"<\/p>\n<\/div>\n<\/div>\n

        <\/p>\n

        \n

        4. UX Strategy & Key Decisions<\/h4>\n\n\n\n\n\n\n\n
        Challenge<\/th>\nApproach<\/th>\nSolution<\/th>\n<\/tr>\n
        Multi-session AI conversations<\/strong><\/td>\nEvaluated navigation models<\/td>\nDrawer-based session management with project grouping (more scalable for power users)<\/td>\n<\/tr>\n
        Building trust in AI outputs<\/strong><\/td>\nTrust pyramid framework<\/td>\nInline citations with bbox-level source highlighting and confidence indicators<\/td>\n<\/tr>\n
        Research vs. Reading<\/strong><\/td>\nRole-based views<\/td>\nEditor (full workspace) vs. Reader (clean consumption with interaction)<\/td>\n<\/tr>\n
        100+ documents in context<\/strong><\/td>\nInformation architecture<\/td>\nProject\/section hierarchy, smart search, prepared context<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

        Core Design Principles<\/strong><\/p>\n

        \n
        \n
        1. Dual-Pane Layout<\/h5>\n

        Left is AI conversation, right is original PDF.
        \nWhy:<\/strong> Builds trust by allowing users to verify AI claims instantly.<\/p>\n<\/div>\n

        \n
        2. Citation-First AI<\/h5>\n

        Every insight links to exact position in source (bbox-level highlighting).
        \nWhy:<\/strong> Eliminates “AI hallucination” anxiety.<\/p>\n<\/div>\n<\/div>\n<\/div>\n

        <\/p>\n

        \n

        5. Final Solution & Interface Design<\/h4>\n

        Translating research and strategy into a production-ready workspace optimized for trust and deep analysis.<\/p>\n

        1. Onboarding Experience<\/strong>
        \nIntroducing the workspace value proposition to new professional users.<\/p>\n

        \"\" \"\"<\/div>\n


        \n2. Core Workspace (Dual-Pane Layout)<\/strong>
        \nThe primary work environment featuring the conversation on the left and the source document on the right.<\/p>\n

        \"\" \"\" \"\"<\/p>\n


        \n3. Citation Navigation (The “Aha” Moment)<\/strong>
        \nWhen a user clicks an AI citation, the right pane automatically scrolls to and highlights the exact bounding box in the original PDF.<\/p>\n

        \"\"<\/div>\n


        \n4. Document Management & Intelligence<\/strong>
        \nHandling uploads, project organization, and AI-generated populated data reports.<\/p>\n

        \"\"<\/p>\n<\/div>\n

        <\/p>\n

        \n

        6. Technical Architecture & Execution<\/h4>\n

        Frontend State Management I Defined:<\/strong><\/p>\n\n\n\n\n\n
        Service<\/th>\nUX Touchpoint<\/th>\n<\/tr>\n
        Agentic Agent Service<\/strong><\/td>\nMulti-agent AI with streaming responses, chat UI, loading states, error handling.<\/td>\n<\/tr>\n
        Data Ingestion Service<\/strong><\/td>\nOCR, markdown conversion, citation mapping, upload flows, processing feedback.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

        Product Roadmap<\/strong><\/p>\n

          \n
        • Internal MVP (Q3 2026):<\/strong> Core workspace UI, chat, PDF viewer, citation navigation<\/li>\n
        • Public MVP (Q4 2026):<\/strong> Landing experience, onboarding, report views<\/li>\n
        • V1.0 (Q1 2027):<\/strong> Session management, agent customization, publishing flow<\/li>\n<\/ul>\n<\/div>\n

          <\/p>\n

          \n

          7. Business Impact & Outcomes<\/h4>\n

          Designed to drive subscription revenue by solving high-value workflow problems for financial professionals.<\/p>\n

            \n
          • Target MAU (6 months):<\/strong> 1,000+<\/li>\n
          • Monthly Rev Target:<\/strong> 1B VND<\/li>\n
          • Feature Adoption:<\/strong> 80%+<\/li>\n
          • Target NPS:<\/strong> \u2265 50<\/li>\n
          • Revenue model:<\/strong> Subscription (Pro tier ~$40\u201360\/month for professionals).<\/li>\n
          • Business goal:<\/strong> 1,000 expert users \u00d7 ~1M VND\/month = 1B VND\/month revenue.<\/li>\n<\/ul>\n<\/div>\n

            <\/p>\n

            \n

            8. Behind the Scenes: Constraints & Lessons<\/h4>\n
            \n
            \n

            Working Conditions<\/strong><\/p>\n

              \n
            • 4-day design sprints:<\/strong> Required ruthless prioritization. Shipped core workspace first, layered details later.<\/li>\n
            • Parallel Work:<\/strong> Managed 2-3 concurrent projects, leading to heavy context switching.<\/li>\n
            • Clear Handoff:<\/strong> Delivered primary screens; PM annotated interaction details in a second pass.<\/li>\n<\/ul>\n<\/div>\n
              \n

              What I’d Do Differently<\/strong><\/p>\n

                \n
              • Advocate for UX Runway:<\/strong> Push for a minimum 2-week UX runway per feature instead of 4 days.<\/li>\n
              • AI Architecture:<\/strong> Involve UX in AI model discussions earlier. Decisions about context limits and hallucination fallbacks happen before Figma.<\/li>\n<\/ul>\n<\/div>\n<\/div>\n
                \n

                The ultimate takeaway: Trust is the product.<\/strong> In AI-for-finance, trustworthiness > features. Every design decision, from bbox citations to dual-pane layouts, must center on verifiability.<\/p><\/blockquote>\n<\/div>\n<\/div>\n<\/div>\n","protected":false},"excerpt":{"rendered":"

                An AI-powered financial research platform. Discover how I single handedly designed a citation-first, dual-pane interface to build trust and solve AI hallucination for professional investors<\/p>\n","protected":false},"author":1,"featured_media":836,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[4],"tags":[],"_links":{"self":[{"href":"https:\/\/meta.ngochn.com\/index.php\/wp-json\/wp\/v2\/posts\/810"}],"collection":[{"href":"https:\/\/meta.ngochn.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/meta.ngochn.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/meta.ngochn.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/meta.ngochn.com\/index.php\/wp-json\/wp\/v2\/comments?post=810"}],"version-history":[{"count":25,"href":"https:\/\/meta.ngochn.com\/index.php\/wp-json\/wp\/v2\/posts\/810\/revisions"}],"predecessor-version":[{"id":843,"href":"https:\/\/meta.ngochn.com\/index.php\/wp-json\/wp\/v2\/posts\/810\/revisions\/843"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/meta.ngochn.com\/index.php\/wp-json\/wp\/v2\/media\/836"}],"wp:attachment":[{"href":"https:\/\/meta.ngochn.com\/index.php\/wp-json\/wp\/v2\/media?parent=810"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/meta.ngochn.com\/index.php\/wp-json\/wp\/v2\/categories?post=810"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/meta.ngochn.com\/index.php\/wp-json\/wp\/v2\/tags?post=810"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}