ForgeDesk

Run platform work and support escalations in one clear command center.

ForgeDesk is a Studio Admin workspace for platform owners, super admins, DevOps engineers, and L3 support teams to manage projects, issues, and B2B escalation tickets in one system. It separates internal engineering work from tenant-originated support while reducing terminology confusion, improving triage speed, and making cross-tenant operations easier to audit and resolve.

Business Goals

  • Reduce average time to first response on B2B escalation tickets by 40% within 90 days.
  • Increase issue resolution throughput per engineer by 25% within one quarter by improving prioritization and screen density.
  • Cut misclassified work items and terminology-related support errors by 80% within 60 days after rollout.
  • Improve SLA compliance for urgent operational issues to 95% or higher within 6 months.
  • Increase weekly active usage of the admin task workspace by 30% within 120 days.

User Goals

  • Help admins quickly distinguish projects, internal issues, and tenant escalations without mental overhead.
  • Let engineers triage, assign, and update work items from a dense, keyboard-friendly interface.
  • Make cross-tenant support tickets visible immediately with clear urgency and ownership.
  • Allow project leads to track roadmap progress automatically from child issue completion.
  • Provide fast access to logs, attachments, comments, and audit history from one place.

Non-Goals

  • Not a full customer-facing helpdesk for end tenants.
  • Not a billing, CRM, or asset management system.
  • Not a generic document editor beyond project specs and issue context.
  • Not a replacement for source-code issue trackers like Jira or GitHub Issues for engineering teams outside Studio Admin scope.

Rafi, 39, Platform Super Admin - Rafi manages cross-tenant platform stability, roadmap execution, and incident coordination. He needs a single place to see strategic projects, urgent platform issues, and escalation tickets without confusing labels or layouts.

Rafi, 39, Platform Super Admin

  • As a super admin, I want to create a project with milestones and linked issues, so that I can track delivery against platform initiatives.
  • As a super admin, I want to see urgent and unresolved escalations at a glance, so that I can respond before SLA breaches.
  • As a super admin, I want consistent labels for projects, issues, and tickets, so that I do not waste time interpreting the interface.

Dina, 32, DevOps Engineer - Dina handles incidents, deployment tasks, and infrastructure tuning. She works quickly across many tickets and needs keyboard-driven triage, compact views, and reliable filtering.

Dina, 32, DevOps Engineer

  • As a DevOps engineer, I want a dense list view with filters and keyboard shortcuts, so that I can process many issues without scrolling constantly.
  • As a DevOps engineer, I want to attach logs and screenshots to an issue, so that I can document the failure context.
  • As a DevOps engineer, I want status changes and priority updates to be audited automatically, so that operational history is traceable.

Mika, 28, L3 Support Lead - Mika owns escalations from ISP tenants and coordinates with platform engineers. He needs clear inbox visibility, strong assignment workflow, and notification handling for new tenant-originated issues.

Mika, 28, L3 Support Lead

  • As an L3 support lead, I want B2B escalation tickets to appear in a dedicated inbox, so that I can separate tenant issues from internal work.
  • As an L3 support lead, I want unread or unassigned escalations to be highlighted, so that no tenant request is missed.
  • As an L3 support lead, I want to comment with structured updates and close the loop to the tenant, so that communication stays clear and professional.

Terminology and taxonomy · High priority

  • The product must use one consistent vocabulary across navigation, list views, buttons, and entity types so users always understand what they are working on.
  • Use Projects for strategic initiatives and Issues for executable work items.
  • Support issue subtypes such as Internal Task, Bug, and B2B Escalation without changing the top-level noun.
  • Show scope badges on every row, such as Platform Internal and B2B Escalation.
  • Replace any mixed labels like task, ticket, and issue with a single controlled taxonomy.
  • Provide a migration map for legacy labels so existing records still render correctly.

Workspace navigation and filtering · High priority

  • The sidebar must organize work by workspace, saved views, and personal filters to support high-volume operational triage.
  • Sidebar must include Projects, Internal Issues, B2B Escalations, All Issues, Active, Overdue, and No Assignee.
  • Each view must preserve its filter state in the URL and be shareable.
  • Scope filters must clearly separate platform work from tenant-originated work.
  • Counters must update in real time for unread, unassigned, overdue, and urgent items.
  • Users must be able to save custom filters as personal views.

Issue and project management · High priority

  • Users must be able to create, update, assign, and resolve projects and issues with linked relationships and status tracking.
  • Support creation of Projects with title, owner, target date, spec editor, and linked issues.
  • Support Issue fields for title, description, scope, type, priority, status, assignee, due date, parent project, and attachments.
  • Automatically calculate project progress based on completion of child issues.
  • Allow issue status transitions including Todo, In Progress, Blocked, Resolved, and Closed.
  • Show validation for required fields and prevent invalid status transitions.

Notifications and auditability · Medium priority

  • Operational changes must be visible immediately and traceable through audit logs and notifications.
  • Emit real-time updates to connected clients when issues are created, updated, assigned, or resolved.
  • Send alerts for urgent or at-risk issues to configured channels such as Telegram, Slack, or WhatsApp.
  • Write every state mutation to an immutable audit log with actor, timestamp, and before/after values.
  • Surface unread escalation indicators in the sidebar and inbox views.
  • Provide comment threads and activity history per issue.

Attachments and integrations · Medium priority

  • Users must be able to attach operational evidence and connect to supporting systems for faster diagnosis.
  • Allow file uploads for screenshots, logs, PCAPs, and spec documents.
  • Integrate object storage for large files with resumable upload support.
  • Provide links or embedded previews for related monitoring, runbook, or incident tools.
  • Expose APIs for external systems to create issues and update statuses.
  • Support SSO-based access control tied to platform roles.

First-Run Workspace Setup

  • User lands on the Studio Admin workspace and sees a clear default split between Projects and Issues.
  • A short welcome panel explains internal work versus B2B escalations in one sentence.
  • User chooses a density mode: Standard or Compact, with Compact recommended for operators.
  • User imports or creates the first project or issue in under 2 minutes using a primary CTA.
  • The system auto-fills the correct workspace scope based on role and source.
  • Time to first useful action target: under 90 seconds for returning users and under 3 minutes for first-time admins.

1. Open Workspace

  • The user opens the Projects and Issues hub and lands on a role-aware default view.
  • Default view should prioritize the user’s most relevant queue, such as My Issues or B2B Escalations.
  • Show scope badges and unread counts immediately.
  • If permissions are limited, hide inaccessible queues rather than showing empty states.

2. Create or Triage Work

  • The user creates a new project or issue, or triages an incoming escalation from the list.
  • Keyboard shortcut C opens the new issue dialog anywhere in the workspace.
  • New project creation should include title, owner, target date, and template selection.
  • Validate required fields inline and prevent submit if scope, title, or priority is missing.

3. Assign and Coordinate

  • The user assigns ownership, adds comments, and links issues to projects for delivery tracking.
  • Allow assignee changes, watchers, and mentions in the activity thread.
  • If an issue is linked to a project, update project progress automatically.
  • If no assignee is available, flag the item as unassigned and surface it in a dedicated view.

4. Resolve and Close

  • The user resolves the issue, captures the outcome, and closes the loop with visible history.
  • Support resolution statuses with a required resolution note for urgent or escalated issues.
  • Trigger notifications on status changes to resolved or closed.
  • Write an audit event for every status and priority mutation.

5. Review and Optimize

  • The user adjusts filters, density, and overview visibility to keep the workspace fast under heavy load.
  • Compact mode should collapse KPI cards into a thin summary strip.
  • Preserve filter state and list scroll position during refresh.
  • Show empty, loading, and error states without breaking keyboard flow.

Power-User and Edge Features

  • Compact density mode with a collapsible overview panel for high-volume triage.
  • Saved views for urgent internal bugs, overdue escalations, and no-assignee queues.
  • Bulk actions for assignment, priority changes, and status updates.
  • Real-time unread badges for incoming B2B escalations.
  • Attachment upload with large-file handling and preview support.
  • Automated linking of issues to projects and progress rollups.
  • Audit history diff view for compliance and debugging.

Visual and Interaction Principles

  • Use a clear visual hierarchy that separates navigation, summary stats, and actionable lists.
  • Avoid mixed terminology in headers, buttons, and row labels.
  • Make scope badges color-coded but accessible with text labels and icons.
  • Design for keyboard-first usage with shortcuts for create, search, filter, and status changes.
  • Support at least 20 visible rows in compact mode on a 1440px display.
  • Maintain WCAG 2.1 AA contrast and visible focus states.
  • Keep initial list render under 1.5 seconds and filter changes under 300 milliseconds when possible.

Rafi is the platform owner for FTTH GIS operations. Every morning he opens the old admin portal and spends too much time decoding whether a record is a task, ticket, issue, or project before he can even start triage.

With ForgeDesk, he lands directly in a clean Projects and Issues workspace. B2B escalations are clearly separated from internal platform work, urgent items are highlighted instantly, and he can create, assign, and resolve work with a few keyboard actions. The result is faster response, fewer classification mistakes, and a clearer operating model for the whole platform team.

Over time, the team sees fewer missed escalations, more predictable project progress, and less time lost to navigation and label confusion. ForgeDesk turns the admin portal from a cluttered list into a reliable command center for engineering and support execution.

User-Centric Metrics

  • Median time from incoming B2B escalation to first response below 10 minutes within 90 days.
  • At least 85% of active users can identify the correct work scope on first glance in usability testing.
  • Reduce time spent searching and filtering work items by 30% within 60 days.
  • Achieve 95% completion of required resolution notes on urgent issues.
  • Increase weekly task completion rate per active engineer by 20% within one quarter.

Business Metrics

  • Increase recurring weekly active users in the admin workspace by 30% within 120 days.
  • Reduce SLA breaches on escalation tickets by 25% within 6 months.
  • Improve internal operational throughput by 20% without increasing headcount.
  • Decrease support rework caused by misrouted work items by 80% after terminology standardization.

Technical Metrics

  • 99.9% monthly uptime for the workspace and issue APIs.
  • List and filter interactions should respond in under 300 milliseconds for typical queues.
  • Real-time badge and notification updates should arrive within 5 seconds of event creation.
  • All privilege-sensitive actions must be audit logged with 100% coverage.

Tracking Plan

  • Track issue_created with scope, type, priority, and source.
  • Track issue_assigned with assignee, previous assignee, and queue.
  • Track issue_status_changed with from status, to status, and resolution time.
  • Track project_created with template, owner, and target date.
  • Track project_progress_updated with percent complete and linked issue count.
  • Track density_mode_changed with standard or compact selection.
  • Track escalation_notification_sent with channel and delivery status.

Technical Needs

  • Frontend built with React and TypeScript using a component system such as shadcn/ui or MUI.
  • Backend API with Node.js or NestJS exposing issue, project, comment, and audit endpoints.
  • Real-time delivery using Server-Sent Events or WebSockets for inbox updates and badges.
  • PostgreSQL for relational project, issue, and audit storage.
  • Redis for caching counters, unread state, and rate-limited notification fanout.
  • Object storage such as Amazon S3 or MinIO for attachments and logs.
  • Role-based access control with SSO-friendly session handling and scoped permissions.

Integration Points

  • Slack for urgent operational alerts and escalation notifications.
  • Telegram or WhatsApp gateway for on-call routing alerts.
  • Sentry or similar error monitoring for issue context and incident correlation.
  • PostgreSQL-backed audit service or an internal gateway-audit service for immutable logs.
  • Auth provider such as Keycloak, Auth0, or Azure AD for role-based sign-in.

Data Storage & Privacy

  • Store only necessary personal data for assignees, commenters, and tenant contacts.
  • Separate tenant-originated escalations from internal platform records at the schema or row-permission level.
  • Encrypt data in transit with TLS and sensitive files at rest with managed key encryption.
  • Apply retention rules for audit logs, attachments, and deleted comments according to company policy and local compliance requirements.
  • Support GDPR and CCPA workflows for export, deletion, and access requests where applicable.

Scalability & Performance

  • Use indexed queries for queue filters, assignee views, and status counts.
  • Paginate and virtualize long issue tables to support queues of 10,000 plus records.
  • Cache real-time counts and badge values to avoid expensive aggregate recomputation.
  • Keep attachment upload off the main request path through async object storage handling.

Potential Challenges

  • Terminology drift across legacy records could confuse users; mitigate with a controlled taxonomy migration and UI alias mapping.
  • High-volume queues may slow the interface; mitigate with virtualization, indexing, and compact mode.
  • Real-time notifications may become noisy; mitigate with user preferences, severity thresholds, and channel routing rules.
  • Cross-tenant permission leaks are a serious risk; mitigate with strict RBAC, tenant scoping tests, and security review.
  • Audit trail growth may increase storage costs; mitigate with compression, lifecycle policies, and partitioned audit tables.

Team & resourcing - Small cross-functional team - 2 frontend engineers, 2 backend engineers, 1 designer, part-time PM, and shared DevOps support.

Phase 1: MVP Workspace · Weeks 1-4

  • Projects and Issues navigation
  • Unified taxonomy for projects and issues
  • Basic list view with filters and search
  • Create and edit issue dialog
  • Role-based access control and scope badges
  • Compact mode toggle prototype

Phase 2: Operational Workflow · Weeks 5-8

  • Issue assignment and status transitions
  • Project creation with linked issues
  • Progress rollup from child issue completion
  • Comments and activity timeline
  • Attachment upload to S3 or MinIO
  • Keyboard shortcuts for create, search, and triage

Phase 3: Real-Time Support and Audit · Weeks 9-12

  • SSE or WebSocket live updates
  • Unread B2B escalation inbox
  • Audit log capture for every mutation
  • Urgent alert routing to Slack, Telegram, or WhatsApp
  • Saved views and personal filters
  • Performance tuning for large queues

Phase 4: Scale and Polish · Weeks 13-16

  • Advanced permissions and tenant isolation hardening
  • Usability refinements for dense list workflows
  • Analytics dashboard for response and resolution metrics
  • Bulk actions and admin power tools
  • Accessibility audit and WCAG fixes
  • Migration tools for legacy task and ticket labels

Paste this into Cursor, Bolt, Lovable, or v0 to start building.

Build a web app called ForgeDesk for a Studio Admin portal that manages platform Projects, internal Issues, and B2B escalation tickets in one workspace. Use a modern stack: Next.js 14, React, TypeScript, Tailwind, shadcn/ui, PostgreSQL, Prisma, Redis, and Server-Sent Events for live updates.

Core requirements:
1. Navigation with Projects, Internal Issues, B2B Escalations, All Issues, Active, Overdue, No Assignee, and saved views.
2. Unified taxonomy only: Project, Issue, scope badges, status, priority, assignee, due date, comments, activity log, attachments.
3. Dense list view with search, filters, bulk actions, compact mode toggle, and keyboard shortcut C to create an issue.
4. Project detail page with issue list and automatic progress rollup from resolved child issues.
5. B2B escalation inbox with unread badge, real-time updates, and notification-ready events.
6. Issue detail drawer/page with description, comments, history, attachments, and status transitions.
7. Audit logging for every create/update/assign/status change.
8. Role-based access control for Platform Owner, Super Admin, DevOps Engineer, and L3 Support Lead.

Data model:
User, Role, Project, Issue, Comment, Attachment, ActivityEvent, NotificationPreference, SavedView, Tenant, EscalationSource.
Project has many Issues. Issue belongs to optional Project, has scope, type, priority, status, assignee, creator, due date, unread flag, and optional tenant source.
ActivityEvent stores immutable before/after changes.
Attachments upload to object storage.

Primary screens:
Workspace dashboard with compact stats strip and list/table of issues.
Projects index and project detail page.
Issue detail page or right-side drawer.
B2B Escalations inbox.
Settings for density mode, notification preferences, and saved filters.

Behavior:
Preserve filter state in URL. Virtualize long lists. Show loading, empty, and permission-denied states. Validate required fields inline. Add optimistic updates for status and assignment. Make the UI accessible with keyboard navigation and AA contrast.

Implement seed data for sample projects, internal bugs, and tenant escalation tickets. Provide realistic mock API routes or server actions so the app runs end-to-end.

Business Idea

# 🏛️ Analisa & Rekomendasi Strategis: Arsitektur, Logika Operasional, dan UI/UX Task Gateway (Studio-Admin) **Dokumen**: Task & Project Management Gateway — Architectural Review & UX Blueprint **Konteks**: Platform Super Admin Portal (`apps/studio-admin`) **Target Pengguna**: Platform Owner, Super Admin, Core DevOps / Infrastructure Engineers, L3 Escalation Support **Lokasi Berkas**: `/opt/project5/docs/06_Archive_Dev_History/05_Plans_Roadmap/TASK_GATEWAY_ARCHITECTURE_AND_UX_RECOMMENDATION.md` --- ## Executive Summary Modul **Tasks & Tickets** (atau *Task Gateway*) pada portal **Studio Admin (`studio-admin`)** memegang peranan sentral sebagai **Pusat Komando Operasional & Rekayasa Platform FTTH GIS**. Berbeda dengan portal tenant (`studio-tenant`) yang berorientasi pada operasional ISP tingkat akhir (pelanggan, ODP, ONT, billing retail), Studio Admin bertindak sebagai payung tata kelola infrastruktur multi-tenant, roadmap platform, dan resolusi eskalasi B2B. Dokumen ini menyajikan audit menyeluruh dari perspektif **Senior Engineer / Product Architect & UI/UX Designer**, menguraikan kekuatan (*strengths*), kelemahan (*flaws*), benturan taksonomi (*terminology clash*), serta cetak biru (*blueprint*) alur kerja dan rekomendasi desain terbaik untuk jangka panjang. --- ## 1. 🎯 Perbedaan Fundamental: Studio-Admin vs Studio-Tenant Sebelum menata taksonomi dan alur kerja, kita harus menetapkan batas tanggung jawab (*separation of concerns*) yang tegas: | Dimensi | 🏢 Studio-Admin (Portal Utama Platform) | 🌐 Studio-Tenant (Portal ISP Mitra) | | :--- | :--- | :--- | | **Audiens Utama** | Super Admin, Platform Architect, DevOps Engineer, L3 Support K2NET | Owner ISP, Helpdesk ISP, Teknisi Lapangan ISP | | **Skala Data** | Cross-Tenant (Semua Tenant + Platform Internal) | Single-Tenant (Hanya data tenant bersangkutan) | | **Entitas Payung (Project)** | Inisiatif Arsitektur, Refactoring Gateway, Fiber Backbone Rollout, Disaster Recovery Sync | Proyek Penarikan Kabel Feeder Lokal, Pembangunan Cluster Perumahan Baru | | **Unit Tugas (Issue/Ticket)** | Bug Microservice, Insiden Kong Ingress, Rate Limit Tuning, Database Optimization | Gangguan Kabel Putus, ONT Red Light, Pemasangan Baru (PSB), Tiket Pelanggan | | **Eskalasi B2B** | **Penerima Eskalasi** (Menyelesaikan isu infrastruktur dari ISP) | **Pengirim Eskalasi** (Meminta bantuan L3 ke tim K2NET) | --- ## 2. 🔍 Audit Kritis UI/UX & Logika Operasional Saat Ini Berdasarkan tinjauan visual pada antarmuka `/tasks` saat ini: ``` ┌──────────────────────────────────────────────────────────────────────────────────────────────────┐ │ [Logo] FTTH GIS K2NET › Tasks [Search ⌘K] [Ask AI Ctrl+J] [UserNav] │ ├─────────────────┬────────────────────────────────────────────────────────────────────────────────┤ │ Tasks & Tickets │ 📋 All Issues (Internal + B2B Inbox) [+ New Issue C]│ ├─────────────────┤ 1 Active Tasks / 0 Urgent / 0 Resolved Today / 2 Total │ │ WORKSPACE ├────────────────────────────────────────────────────────────────────────────────┤ │ 📁 Projects │ ┌──────────────────────┐ ┌──────────────────────┐ ┌────────────────────────┐ │ │ 📑 Internal Task│ │ ACTIVE TASKS │ │ URGENT │ │ RESOLVED TODAY │ │ │ 🏢 B2B Tickets │ │ 1 │ │ 0 │ │ 0 │ │ │ │ └──────────────────────┘ └──────────────────────┘ └────────────────────────┘ │ │ VIEWS ├────────────────────────────────────────────────────────────────────────────────┤ │ 🗃️ All Issues │ [🔍 Filter by title...] [Status ▾] [Priority ▾] [Scope ▾] [Display] [Refresh] │ │ ⚡ Active ├────────────────────────────────────────────────────────────────────────────────┤ │ ⏰ Overdue │ [ ] Title Scope Type Priority Status Assignee Due │ │ 👤 No Assignee │ [ ] Project 1 PRJ-002 [Internal] PROJECT LOW Todo Admin 25A │ │ │ [ ] Setup S3 Sync [Internal] TICKET HIGH In Prog DevOps 24A │ └─────────────────┴────────────────────────────────────────────────────────────────────────────────┘ ``` ### A. Evaluasi Taksonomi & Penamaan (Terminology Clash) - **Temuan Masalah**: - Modul di sidebar diberi label **`Tasks & Tickets`**. - Header utama di atas tabel tertulis **`All Issues`**. - Tombol CTA tertulis **`+ New issue`** (dengan shortcut `C`). - Kartu KPI di atas bertuliskan **`ACTIVE TASKS`**. - Kolom tabel menampilkan **`Type: PROJECT`** vs **`Type: TICKET`**. - **Dampak UX**: Pengguna mengalami kebingungan mental (*cognitive friction*). Pengguna bertanya-tanya: *"Apakah saya sedang mengelola Task, Ticket, Issue, atau Project?"*. - **Rekomendasi**: Mengadopsi taksonomi standar industri (seperti **Linear** atau **GitHub Enterprise**) yang konsisten di seluruh lapisan UI. --- ### B. Evaluasi Hirarki Navigasi Sidebar Kiri - **Kelebihan (Strengths)**: 1. **Struktur 3-Tier (*Workspace*, *Views*, *Personal*)**: Ini adalah pola UX kelas dunia (mengikuti standar Linear.app) yang mempermudah navigasi cepat berdasarkan konteks kerja. 2. **Pemisahan Scope (*Internal Tasks* vs *B2B Mitra Tickets*)**: Sangat krusial bagi Super Admin untuk membedakan pekerjaan internal platform dari permintaan bantuan mitra ISP. - **Kelemahan & Ambiguitas (Weaknesses)**: 1. **Duplikasi Konteks *All Issues* vs *Internal Tasks***: Mengklik *All Issues* menampilkan gabungan, sementara *Internal Tasks* memfilter `PLATFORM_INTERNAL`. Perlu visual badge yang memperjelas scope tiket (misal badge warna biru `Platform Internal` vs ungu `B2B Escalation`). 2. **Badge Counter Belum Dinamis**: Menu *B2B Mitra Tickets* harus menampilkan badge merah/amber menyala jika ada tiket masuk baru dari tenant ISP yang belum di-respon (*unread/unassigned*). --- ### C. Evaluasi Kepadatan Visual (*Visual Density & Screen Real Estate*) - **Temuan Masalah**: - **KPI Cards Besar (*Active Tasks, Urgent, Resolved*)** memakan tinggi sekitar 140px di bagian atas layar. - Untuk engineer dan DevOps yang bekerja dengan daftar 50+ tiket per hari, area vertikal ini membuat tabel terdorong ke bawah (*below the fold*), mengharuskan engineer melakukan scrolling berulang kali. - **Rekomendasi UX**: - Menyediakan **Mode Compact Stats Strip** (1 baris ringkas 32px di header) atau tombol *Collapse Overview Panel* agar engineer bisa memaksimalkan ruang layar untuk tabel tiket dan kanban board. --- ## 3. 📐 Blueprint Arsitektur & Hirarki Entitas yang Ideal Untuk menciptakan sistem manajemen proyek dan rekayasa platform yang solid, berikut struktur relasi entitas (*Entity-Relationship Model*) yang direkomendasikan: ``` ┌──────────────────────────────────────────────────────────────────────────────────┐ │ LEVEL 1: INITIATIVES & ROADMAPS │ │ (Projects & Plans — /tasks/projects) │ │ - Karakteristik : Payung besar inisiatif strategis platform (Lead, Target Date) │ │ - Dokumen : Spesifikasi Teknis TipTap Markdown (Tech Spec, Arch, SLA) │ │ - Progress : Otomatis terhitung (% akumulasi issue anak yang Resolved) │ └────────────────────────────────────────┬─────────────────────────────────────────┘ │ │ 1-to-Many Relasi (parentTaskId / obsidianRef) ▼ ┌──────────────────────────────────────────────────────────────────────────────────┐ │ LEVEL 2: WORK ITEMS & ISSUES │ │ (Tasks & Issues — /tasks) │ │ │ │ ┌────────────────────────────────────────┐ ┌───────────────────────────────┐ │ │ │ A. PLATFORM INTERNAL ISSUES │ │ B. B2B MITRA ESCALATIONS │ │ │ │ (Scope: PLATFORM_INTERNAL) │ │ (Scope: TENANT_TO_PLATFORM) │ │ │ │ - Bug fixing gateway/core API │ │ - Tiket eskalasi dari ISP │ │ │ │ - Task deployment / migration │ │ - Permintaan routing khusus │ │ │ │ - Optimasi database & query tuning │ │ - Kendala payment / OLT poller│ │ │ └────────────────────────────────────────┘ └───────────────────────────────┘ │ └────────────────────────────────────────┬─────────────────────────────────────────┘ │ │ 1-to-Many ▼ ┌──────────────────────────────────────────────────────────────────────────────────┐ │ LEVEL 3: SUB-ISSUES & CHECKLISTS │ │ (TaskSubIssuesSection / Milestones Checklist) │ │ - Unit eksekusi mikro per engineer (Checklist, PR Review, Docker build test) │ └──────────────────────────────────────────────────────────────────────────────────┘ ``` --- ## 4. 🔄 Alur Operasional Baku (End-to-End Operational Workflows) ### Alur 1: Inisiatif Platform Terencana (Top-Down Engineering) ``` 1. Super Admin / Lead Architect membuka `/tasks/projects`. 2. Klik "+ New Project" -> Pilih Template "Tech Spec" -> Tulis Dokumen Desain di TipTap. 3. Tentukan Target Date (misal: 30 September 2026) dan Engineering Lead. 4. Masuk ke tab "Issues" di dalam Project -> Tambahkan 5-10 Issue teknis terinci. 5. Engineer menerima tiket di menu `/tasks` ("My Issues") -> Mengubah status ke IN_PROGRESS. 6. Saat seluruh Issue selesai (RESOLVED), progress Project otomatis menjadi 100%. ``` ### Alur 2: Penanganan Insiden & Bug Harian (Bottom-Up Incident Response) ``` 1. Engineer menemukan anomali (misal: Redis Cache Spike di Poller Gateway). 2. Tekan shortcut 'C' di mana saja pada `/tasks` -> Buka "New Issue Dialog". 3. Pilih Priority: URGENT / HIGH -> Masukkan deskripsi log insiden. 4. Jika insiden terkait dengan inisiatif yang ada, kaitkan dengan Project terkait (opsional). 5. Tim menyelesaikan perbaikan -> Tandai status RESOLVED -> Resolved Today counter bertambah. ``` ### Alur 3: Penanganan Tiket Eskalasi B2B dari Mitra ISP (Cross-Tenant Support) ``` 1. Admin ISP Mitra mengalami kendala OLT di studio-tenant -> Klik "Eskalasi ke K2NET Support". 2. Backend membuat tiket dengan `scope: TENANT_TO_PLATFORM`. 3. Notifikasi realtime berdering via SSE di Studio Admin -> Badge merah di "B2B Mitra Tickets". 4. L3 Support K2NET membuka tiket -> Berdiskusi di tab Activity/Comments. 5. Masalah teratasi -> Tiket ditutup -> Tenant ISP menerima notifikasi instan. ``` --- ## 5. 💡 Rekomendasi Aksi & Roadmap Penyempurnaan UI/UX ### Rekomendasi 1: Standardisasi Taksonomi (Unified Terminology) - **Nama Modul Utama**: Ganti dari `Tasks & Tickets` menjadi **`Projects & Issues`** (atau **`Engineering Hub`**). - **Entitas Level 1**: **`Projects`** (Platform Initiatives, Roadmaps, Milestones). - **Entitas Level 2**: **`Issues`** (Sub-kategori: *Internal Task*, *Bug*, *B2B Support Ticket*). - **Tombol CTA**: **`+ New Issue`** (Shortcut: `C`) dan **`+ New Project`** (di halaman Projects). ### Rekomendasi 2: Penambahan Toggle Kepadatan Layar (*Density Modes*) - Tambahkan switch pada toolbar tabel: - **Standard View**: Menampilkan KPI cards besar di atas + tabel di bawah. - **Dense / Productivity View**: Menyembunyikan KPI cards besar, hanya menampilkan statistik 1 baris di header, sehingga 20+ baris tiket dapat tampil penuh di layar pertama. ### Rekomendasi 3: Integrasi Otomatisasi Antar-Gateway 1. **Gateway-Audit (`gateway-audit:5009`)**: - Setiap mutasi tiket (perubahan prioritas, assignee, status) otomatis mengalir ke log audit platform (`logGroup: OPERATIONS`). 2. **Gateway-Notification (`notification-gateway:5001`)**: - Kirim alert Telegram / WhatsApp otomatis ke channel DevOps jika ada tiket berstatus `URGENT` atau `At Risk`. 3. **Gateway-Storage & MinIO**: - Dukungan unggah lampiran log file, file Wireshark PCAP, atau tangkapan layar topologi GIS langsung ke bucket S3. --- ## 6. 🏆 Kesimpulan & Dampak bagi K2NET Platform Dengan mengimplementasikan rekomendasi di atas: 1. **Kejelasan Operasional**: Engineer dan Super Admin memiliki batasan kerja yang jelas antara merancang proyek strategis (*Projects*) dan menyelesaikan pekerjaan teknis (*Issues*). 2. **Kecepatan Kerja (Speed & Ergonomics)**: Antarmuka yang bersih, padat, dan bebas benturan terminologi akan mempercepat siklus *triage* bug, penyelesaian insiden, dan penanganan eskalasi tenant. 3. **Standar Kelas Enterprise**: Menjadikan portal K2NET FTTH GIS setara dengan platform SaaS engineering kelas dunia seperti Cloudflare Dashboard, Supabase Studio, dan Linear.

Make My PRD

Design by The Resonance | Powered by GPC – The AI Transformation Company

    PRD: # 🏛️ Analisa & Rekomendasi Strategis: Arsitektur