Agency operations platform

All your team needs, on one screen

Replace scattered Excel, Zalo and Drive with a single place for every client project.

Kanban & content calendar

Drag posts through every production step, a weekly / monthly publishing calendar, and a two-step “Deleted” column so nothing is lost by mistake.

Tasks & appointments

Assignees, collaborators, deadlines, material requests to the client, and appointments synced with Google Calendar.

Weekly KPIs

Targets per channel and campaign; progress updates automatically from synced metrics.

Client portal

Clients view, approve and comment right on each post; internal notes always stay hidden.

Agency finance

Profit and loss on cash actually in and out, an income-expense ledger, team pay based on delivered work, payments and printed payslips.

Contracts & VietQR payments

Service contracts from templates (Vietnamese / bilingual), signed electronically by the client, printable as PDF. Payment terms per month / per deliverable / per package, gentle client reminders, and an automatic VietQR code carrying the transfer note.

Streaks & Hall of fame

Daily check-ins, a shared team streak, a leaderboard and automatic Bronze · Silver · Gold badges.

Granular permissions

Permissions per feature × action (view, create, edit, delete, approve, export), custom roles, and separate data for each project.

Instant & installable app

Realtime updates over WebSocket, push notifications to your phone, and install to the home screen like an app.

Excel plan import

Upload an existing plan file — the system builds the publishing calendar, the content pillars and the task list for you.

Activity history

Every post and every task keeps a log of who changed what — including changes AI made on your behalf.

Auto-publishing Professional plan

Drag a card into the “Scheduled” column on the Kanban — at the exact time you set, the system posts the caption, hashtags and design to Facebook, Instagram, Threads, TikTok and YouTube, pulls the post link back onto the card, moves it to “Published” and writes the log.

Enabled per channel and per post, off by default: nothing goes out unless you switched that channel on yourself and ticked that exact post. TikTok and YouTube currently publish in private mode only, until both platforms finish reviewing the app.

For performance marketers

Clear targets, early alerts, per-post attribution

Metrics sync automatically every day; ratios are computed at read time, so KPIs always add up. AI and alerts follow the targets you set.

◎

Target ROAS · CPA · CTR

Set a max CPA, a min ROAS, CTR and a monthly budget for each Google Ads account — with alerts when CPA runs far past target or ROAS falls below it.

⚡

Budget pacing

Projected monthly spend above budget is flagged by severity; unusually slow spending mid-month is flagged too.

⌕

Search terms

The 10 costliest terms, terms with spend and 0 conversions, disapproved ads, low Quality Score keywords — on one screen.

⇄

UTM & attribution

Every post gets its own UTM-tagged link per platform and campaign; GA4 sessions and revenue go back to the post that actually brought the visitor.

△

Early alerts

KPIs off pace, posts behind schedule, engagement dropping sharply, content pillars out of balance, channel sync errors — flagged before the client complains.

◷

Forecasts & critique

A brand funnel, actions ranked by revenue impact, result forecasts and plan critique with an offer check.

Connected data sources

  • Facebook
  • Instagram
  • Threads
  • TikTok
  • Google Ads
  • Meta Ads
  • GA4
  • Search Console
  • Shopify
  • WooCommerce
  • Haravan
  • Sapo

Cloud infrastructure

Runs on a global edge network, with zero physical servers

The app, the databases, file storage, queues, semantic memory and the AI GPUs all sit on Google + Amazon + Cloudflare cloud infrastructure — an edge network spread across more than 300 cities — wired straight into the Google Cloud, Meta, TikTok and Shopify APIs. There is no private server to crash, overload or wait on for maintenance.

  • 300+cities with an edge location running the app — always closest to the user
  • 0physical servers — no failed drives, no overload, no waiting for maintenance
  • 30 dayspoint-in-time data recovery, automatic backups
  • 5 minmetric sync & cost monitoring cadence, running 24/7
⚡

Near-zero latency

Code runs at the edge location closest to you instead of on a distant server; heavy processing is placed next to the data automatically. Every change is pushed instantly to open tabs over WebSocket — no page reloads, no constant polling.

◈

Cloud GPUs for AI, many models

AI runs on Google + Amazon + Cloudflare GPUs with leading language, image and voice models. Each job goes to the model that suits it best, with a backup model when one is overloaded — AI never stops.

⛨

Data safe, never lost

Databases are backed up automatically, with point-in-time recovery over 30 days; files sit in high-durability private storage and are served only through permission-checked APIs. Zero Trust Access stands in front of every request, and channel tokens are AES-GCM encrypted.

⇄

Real-time processing, on cadence

Background job queues, semantic memory and a global cache keep heavy work off the user path: you click and get a response, AI handles the long part behind the scenes and reports back the moment it is done. Cloud spend is measured every 5 minutes, with a daily cap.

Tech & infrastructure partners

  • Google Cloud
  • Amazon Web Services
  • Cloudflare
  • D1
  • R2
  • Durable Objects
  • Queues
  • Workers AI
  • Browser Rendering
  • Zero Trust Access
  • Google Cloud APIs
  • Meta Graph API
  • Shopify API
  • MCP · Claude / ChatGPT