
A 2-Person Team With 10+ Jobs Running Themselves Daily: Javis OS Replaces Repetitive Work, the Owner Just Reads Reports
Vũ Khôi
Tác giả tại Wotech Electronics Blog

A 2-Person Team With 10+ Jobs Running Themselves Daily: Javis OS Replaces Repetitive Work, the Owner Just Reads Reports
Our team has exactly 2 people. But every day, more than 10 jobs run themselves, even while both of us are asleep. That is the sharing of Phạm Quảng Đại in the Sắc Màu Cuộc Sống Xa Quê group (20/11/2026) about his Javis OS — the photo below is the "Recurring jobs" page in Javis OS, where each row is a job that used to need a person sitting down to do it, and now the machine does it and reports back on its own. All figures, workflows, and views in this article follow the post and have not been independently verified by Wotech.
Quick answer: According to the post, Phạm Quảng Đại's Javis OS runs more than 10 jobs a day by itself for a 2-person team: every 5 minutes it scans comments under ad posts, replies and DMs each person (as of the morning of 20/11, the bot had handled 108 comments and DMed 94 people, with no one on duty) and the Threads chatbot auto-replies to comments; at 7 a.m. it pulls yesterday's ad numbers into a sheet, adds tax, then auto-creates a payment slip in the sales software; at 8 a.m. and 8 p.m. it auto-posts content to Instagram and Threads; every 3 hours the assistant reports ad cost and profit to the work group, with a reminder to pay the ads card; on Monday morning it syncs documents from Lark to the brain and auto-renews Instagram and Threads tokens before they expire. Zalo in particular — where Vietnamese customers message the most — has a bot on duty in a class group of 1,000+ students; every 3 minutes it watches the customer group, and when someone tags him with a question, Javis pushes the question to him with a suggested reply ready, and he just approves it; when customers send photos or files via Zalo, the bot reads them all, takes care of them, proactively friends and messages them; every 30 minutes it watches the Zalo bot itself, and reports any problem immediately. There are also batch jobs: posting reels to Facebook, Instagram, and YouTube on a fixed daily schedule, and posting photo-plus-text posts to Facebook and Instagram — he just loads posts into a sheet, and the machine picks the right post, at the right time, on the right channel; reels alone have gone up 30+ posts this way without him pressing post even once.
Javis OS is Phạm Quảng Đại's automation system — per the post, the "Recurring jobs" page in Javis OS lists every repetitive job the machine has taken over from people.
What did he do by hand before?
According to the post, in the early stage of building Javis, he did everything by hand: in the morning, open the ad account and copy numbers into a sheet; create ad-spend payment slips in the sales software; reply to each comment, message each customer; post to each channel one by one; remember token renewal dates so connections would not break. No single job was hard. But together they ate up the whole day, and they were all repetitive.
What does the machine do for him now?
According to the post, the year-round automated schedule looks like this:
| Time | What the machine does, per the post |
|---|---|
| Every 5 minutes | Scans comments under ad posts, replies and DMs each person; Threads chatbot auto-replies to comments |
| 7 a.m. | Pulls yesterday's ad numbers into a sheet, adds tax, auto-creates a payment slip in the sales software |
| 8 a.m. and 8 p.m. | Auto-posts content to Instagram and Threads |
| Every 3 hours | Assistant reports ad cost and profit to the work group, with a reminder to pay the ads card |
| Monday morning | Syncs documents from Lark to the brain, auto-renews Instagram and Threads tokens before expiry |
| Every 3 minutes (Zalo) | Watches the customer group; when someone tags him with a question, pushes it to him with a suggested reply ready for approval |
| Every 30 minutes (Zalo) | Watches the Zalo bot itself, and reports any problem immediately |
The screenshot in the post also shows some other jobs on the list: checking YouTube Shorts status, filling TikTok reel links into a sheet, scanning emails from 3 Skool AI groups to filter on-topic subjects, and pulling viral TikTok reels to Drive.
Apply now: the test-first, hand-over-later principle
According to the post, automation does not mean never breaking. Some days the Facebook connection drops, and the report must state clearly "this source could not be read". Some days the bot's AI hits its quota and stands still for days. So he follows the old principle: test first, and only hand over to the machine once it runs stably. When something breaks, the machine must report it, not stay silent. The lesson he draws: small businesses do not lack ideas; they lack people to do the repetitive daily work. A boss who spends every day copying numbers, replying to messages, and posting has no energy left to think big. For him, AI is worth the money not because it writes well or draws beautifully, but because it does exactly those repetitive jobs, 24/7, never forgetting, never resting. He also invites readers to comment "Auto System" to receive a free repo for building automations like these.
Frequently asked questions
1. What does Javis OS automate? According to the post: scanning and replying to ad comments every 5 minutes, the Threads chatbot, pulling ad numbers into a sheet and creating payment slips at 7 a.m., posting content to Instagram and Threads at 8 a.m. and 8 p.m., reporting ad cost and profit every 3 hours, syncing Lark documents and renewing tokens on Monday morning, staffing the Zalo groups, and posting reels/photo posts in batches.
2. How big is the author's team? According to the post, the team has exactly 2 people, yet more than 10 jobs run themselves every day, even while both of them are asleep.
3. What can the Zalo bot do? According to the post: it staffs a class group of 1,000+ students, answers anyone who asks, and where the documents are unclear, it says plainly that it cannot confirm rather than guessing; it watches the customer group every 3 minutes and pushes questions with a suggested reply for him to approve; it reads photos or files customers send, takes care of them, proactively friends and messages them.
4. Does automation ever break? Per the post itself, yes: some days the Facebook connection drops and the report must state "this source could not be read"; some days the bot's AI hits its quota and stands still for days. His principle: test first, hand over only once it runs stably; when something breaks, the machine must report it, not stay silent.
5. What lesson does the author draw? According to the post: small businesses do not lack ideas; they lack people to do the repetitive daily work; AI is worth the money not because it writes well or draws beautifully, but because it does exactly those repetitive jobs, 24/7, never forgetting, never resting.
Source: post by Phạm Quảng Đại in the Sắc Màu Cuộc Sống Xa Quê group (20/11/2026) about Javis OS. All figures, workflows, and views in this article follow the post and have not been independently verified by Wotech.
Bài viết liên quan

20 tháng 11, 2026 · 6 phut doc
Living Spec: A Controlled Experiment Shows the Spec File Is 54x Cheaper — But a Stale Spec Is a Single Point of Failure
I just made my own AI prove a hot method, and the result made me revise my own understanding. That is the sharing of Cáo Nhăn Nhó in the Muse AI Việt Nam group (19/11/2026) about the Living Spec meth…

20 tháng 11, 2026 · 7 phut doc
Muse Assistant Sues Meta Itself: Locked Out as a Suspected Hacker, Then It Drafted the $75,000 Lawsuit
A small business owner in South Carolina just used Meta's own Muse assistant to sue Meta, and this story is exactly the kind only the agent era could produce. That is the content of a post by Trần Mạ…

20 tháng 11, 2026 · 7 phut doc
Muse Connects to Google Calendar: A Calendar That Reports to AI, So Muse Knows About a Rescheduled Meeting Before You Do
One afternoon, a meeting with a supplier got rescheduled, and before you even opened your calendar to check, Muse already knew. Not because Muse guessed, but because your calendar is connected to it,…