“Do the numbers agree?”
Atelier OS is the business software your whole organisation runs on. One record for every department.
| Order SL-1187 | Workshop | Finishing |
|---|---|---|
| Finance | Advance matched | |
| Stores | Lot 0412 issued | |
| Owner | On time |
RunningExplore Atelier OS
Nizento · Karachi
AI, made to measure for your business.
We study how your organisation really works, then build the software and AI around it. We run our own colleges, restaurants and a fleet of 40 trawlers this way.
For colleges and universities, restaurant groups, fashion and garment houses, and family groups of companies.
- 49 modules running in our own businesses
- Around PKR 5m caught for a client in two months
- 1 week from visit to written proposal
Meet Nizento
Built in businesses we run.
Naveed Mughal and Mustafa Karim run colleges, restaurants, a Subway franchise and a fleet of 40 trawlers. The software on the market made our staff work its way, so we built our own. It runs in them every day.
Edvia CollegeGreenwich Global CollegeWhipped CoffeeFatman BurgersGreenwich PadelPremier Franchise (Subway)Prima Fisheries
About usOur products
Three questions, answered.
“What needs me today?”
Tracer is your assistant, on WhatsApp and email. It tells you what needs you.
- Ask
- Retrieve
- Inspect context
- Review the proposed action
- Advance the workflow
RunningExplore Tracer
Nizento DS
“What if the internet goes down?”
Your business keeps running. Nizento DS is a server in your own office that runs Atelier OS. Orders, cash and stock carry on over your office network, and everything catches up online when the line is back.
RunningExplore Nizento DS
Our Work
Proven where we could watch it every day.
Our Work · SAK Atelier · Client · Fashion house
Around PKR 5m caught in two months.
Orders came in from the online store, WhatsApp, Instagram and the counter, each kept in a different place. We put every order on one board and checked it against the bank.
- 8
- paid orders nobody was making, found on the first check
- 139
- payment screenshots checked against the bank’s own alerts
- 680
- production steps read straight from WhatsApp, none retyped
Nizento’s estimate of SAK Atelier’s first two months, Sept–Oct 2026.
49 of 49
modules in daily use in our own organisations, as of 4 October 2026.
Founders’ statementEach morning
every branch’s cash checked from the photos managers already send.
Whipped Coffee and Fatman BurgersDaily
one morning brief across several companies, used by a founder since September 2026.
Group head officeNizento DS
When the internet goes down, your business keeps running.
Nizento DS is a small server we set up in your own office. Atelier OS runs on it, over your office network. Scroll through the machine.
The machine below is our navy and gold design target, drawn in real time in your browser. The inside is a conceptual view, not an engineering drawing.
Nizento DS
Your system, in your own office.
Nizento DS is a small server we set up in your building. Atelier OS runs on it.
One box for the whole office.
Not one per person. Your staff use the laptops and phones they already have, over your office network.
Inside · a conceptual view
An illustration of the inside, not an engineering drawing.
Your record lives here.
Orders, ledger, payroll, student files. Where each kind of record is kept is agreed with you in writing.
Keeps working when the internet doesn’t.
Atelier OS runs over your office network, so orders, fees and stock carry on during an outage.
Everyday AI, close to your data.
Small jobs, like reading receipts and sorting messages, are being built to run on the box. Harder questions go to a cloud AI, only with data you approve.
Set up, backed up and looked after by us.
Backup power for load-shedding, an encrypted copy off-site every night, and one signed page listing every route out.
Nizento DS
Your system, in your building.
64 GB memory · 2 TB storage · installed and looked after by us
The wider story
One part of a larger system.
Atelier OS runs on the DS, and Tracer reads Atelier OS. Every client chooses: the system online, or in their own office.
Design target · drawn in real time
01
Your system, in your own office.
Nizento DS is a small server we set up in your building. Atelier OS runs on it.
02
Your record lives here.
Orders, ledger, payroll, student files. Where each kind of record is kept is agreed with you in writing.
03
Keeps working when the internet doesn’t.
Atelier OS runs over your office network, so orders, fees and stock carry on during an outage.
04
Everyday AI, close to your data.
Small jobs, like reading receipts and sorting messages, are being built to run on the box. Harder questions go to a cloud AI, only with data you approve.
05
Set up, backed up and looked after by us.
Backup power for load-shedding, an encrypted copy off-site every night, and one signed page listing every route out.
06
Nizento DS. Your system, in your building.
64 GB memory · 2 TB storage · installed and looked after by us
Stills from the real-time model of the design target. The interior is a conceptual view.
- Nizento DS appears as a band of light crosses it. Your system, in your own office: a small server we set up in your building, with Atelier OS running on it.
- The camera approaches the front and enters through a vent. One box for the whole office, not one per person.
- A conceptual view of the inside. Your record lives here: orders, ledger, payroll, student files.
- It keeps working when the internet does not: Atelier OS runs over your office network.
- Points of light travel toward the processor. Everyday AI, close to your data: small jobs are being built to run on the box, and harder questions go to a cloud AI, only with data you approve.
- The lid closes and the Nizento star is revealed. Set up, backed up and looked after by us.
- Nizento DS. Your system, in your building. 64 GB memory · 2 TB storage · installed and looked after by us.
What we build
Pick the part of your day that costs you most.
Each area shows the problem, what changes, and a demo you can try. What already runs in our own businesses is marked Running.
A · Example workflow
Illustration
What needs you, on one page, each morning.
One page for the person at the top: the decisions only they can take, what moved overnight, and an answer to any plain question, with its sources.
The problem
Reports reach the top late, each in its own format. By the time the picture is complete, the moment to act on it has passed.
The few decisions only the owner can take sit in chat groups beside a hundred routine messages.
How it works
- Departments work on one record
- Watchers note what changed overnight
- Tracer ranks what needs you
- You decide with the source in view
- The decision is logged with who and when
Morning brief · Illustrative demo
Morning brief · Tuesday 7 October
- Needs you
- 3 decisions: a second signature on a fabric payment, a till short at Clifton, 14 admission offers
- Money
- Cash checked at all three branches; one short by Rs 3,400
- Customers
- 41 enquiries yesterday; 39 answered within 15 minutes
- Operations
- 2 orders due this week are behind at finishing
Example output · sample data for a fictional organisation
C · Example workflow
Illustration
Ask, check the context, approve the next step.
Tracer is our AI assistant and action layer. It finds what you ask for, shows the records behind it, and prepares the next step for a person to approve.
The problem
What you need for a decision is spread across WhatsApp, email, files and the ERP. Gathering it takes longer than deciding.
Follow-ups depend on someone remembering to send them.
How it works
- Ask
- Retrieve the records
- Inspect the context
- Review the proposed action
- Advance the authorised workflow
Ask, check, approve · Illustrative demo
Proposed action · Clifton till
- Context
- Till Rs 3,400 short last night; counted by Imran at 11:40 pm; two paid-out slips not photographed
- Draft
- Message to Imran asking for the two slips and a recount by noon
- Next step
- Task on the branch board, due 12:00, owner Imran
- Waiting for
- Your approval. Nothing is sent until you approve
Example output · sample data for a fictional organisation
D · Example workflow
Illustration
Every enquiry answered, assigned and followed.
Enquiries from WhatsApp, the website, email and the phone land in one inbox. Each gets an answer from your own price list and policies, an owner and a next step.
The problem
Enquiries wait for hours, or are lost between personal phones and a shared inbox.
Nobody can say how many came in last week, who answered them, or what became of them.
How it works
- An enquiry arrives on any channel
- Answered from your price list and policies
- Assigned to an owner with a next step
- Follow-ups scheduled
- Won or lost, with the reason recorded
Enquiry desk · Illustrative demo
Enquiry desk · yesterday
- Arrived
- 41 enquiries: WhatsApp 26, web form 8, email 5, phone notes 2
- Answered
- 39 within 15 minutes; 2 waiting for a price from the workshop
- Assigned
- Every enquiry has an owner and a next step
- Booked
- 6 measurement visits, 4 campus visits
Example output · sample data for a fictional organisation
E · Example workflow
Illustration
From first enquiry to enrolled student, on one record.
For colleges, universities and school groups: the enquiry desk, the front desk, admissions, attendance and the policies that govern them, on the same record as finance and HR.
The problem
In admission season the enquiry desk is overwhelmed, applicant documents arrive in pieces, and fee payments are confirmed by eye.
Attendance registers and policy binders live apart from the record that pays salaries and collects fees.
How it works
- Enquiry answered
- Application and documents
- Interview booked
- Offer issued
- Fee matched to the bank
- Enrolled; attendance begins
Applicant journey · Illustrative demo
Admissions · evening BBA, January intake
- Applications
- 212 received; 168 complete; 44 waiting for a document
- Interviews
- 36 booked this week; 4 slots free on Saturday
- Offers
- 14 waiting for the principal’s signature
- Fees
- Paid fees matched to the bank each night
Example output · sample data for a fictional organisation
H · Example workflow
Illustration
Every payment matched to the bank. Every purchase approved by limit.
A general ledger for each company, the daily cash check by branch, the nightly bank check, vendor payments with their vouchers, and purchasing approved by amount.
The problem
Cash leaks between the till, the card slips and the bank. Payments in are confirmed by eye from screenshots.
Purchases are approved in chat groups, with no limit per role and no voucher behind the payment.
How it works
- Request raised
- Quotations compared
- Approved by amount: two signatures on money
- Goods received and counted
- Vendor paid with its voucher
- Matched to the bank that night
Nightly bank check · Illustrative demo
Nightly bank check · Monday 6 October
- Matched
- 57 of 59 payments in matched to the bank’s own alerts
- Not matched
- 1 screenshot with no alert; 1 alert with no record
- Vendor payments
- 4 paid, each with its voucher and two signatures
- Cash
- Three branches checked; Clifton short Rs 3,400
Example output · sample data for a fictional organisation
J · Example workflow
Illustration
One board from the first message to delivery.
Every order or job on one board, each piece tracked through its stages, handovers signed by both sides, and a daily check that no confirmed order is missing.
The problem
The floor runs on calls and memory. Orders confirmed on WhatsApp never reach the workshop.
When a piece is late, nobody can say who had it last.
How it works
- Order confirmed on any channel
- Placed on the board
- Moves stage by stage
- Handovers signed by both sides
- Dispatched and tracked
- Daily check: nothing confirmed is missing
Order board · Illustrative demo
Order board · workshop · Tuesday 7 October
- On the board
- 64 orders: 11 cutting, 9 dyeing, 21 with makers, 15 finishing, 8 packed
- Late
- 2 behind at finishing; owners told at 9:00
- Handovers
- 37 yesterday, each signed by both sides
- Daily check
- Every confirmed order is on the board
Example output · sample data for a fictional organisation
No module in this area yet.
How we work
We study before we build, and we stay after.
We do not quote from a brief alone. We begin with one paid visit and one department, and widen from there.
01
Understand
We come to your organisation and learn from the people who do the work: half a day on your floor and four days with your records.
02
Design
Our analysts sit with our engineers and plan where software and AI can help, and how much can be automated. A written proposal follows within 7 days.
03
Build
One department first, fitted to how you already run, and run beside your old way for a month before anything is switched off.
04
Deploy and improve
Department by department, under a steering group. After launch you have a named person, an alert when anything stops, and a monthly review.
Free
A first call, fifteen to thirty minutes.
To beginRs 25,000
The visit: half a day on your floor and four days with your records. Credited if you go ahead within 30 days.
The first paid step7 days
From the visit to a written proposal. Every stage after that is quoted in writing.
What you getQuestions owners ask
Plain answers.
We describe where your data goes precisely, in writing, for your installation. No reassurance in place of fact.
Data and deployment- What does Nizento do?
- We study how your organisation really works, then build the software and AI around it, and stay to support it.
- Who is it for?
- Colleges and universities, restaurant groups, fashion and garment houses, and family groups of companies. Owner-run businesses come through the same door.
- What does it cost to start?
- A first call is free. The visit is Rs 25,000, credited if you go ahead within 30 days. Every stage after that is quoted in writing.
- How long does it take?
- A written proposal within 7 days of the visit. Then one department first, run beside your old way for a month before anything is switched off.
- Do our staff have to change how they work?
- No. The software takes the shape of your work. It prints the sheets your floor already uses and reads the reports your staff already send.
- Does it run anywhere today?
- Yes. 49 modules run every day in the colleges, restaurants and other businesses the founders run, and at a client’s fashion house.
- Where does our data go?
- We tell you in writing before anything is installed: what runs where, what is stored where, and every route out, including our own support access. You decide each line.
- Do we need a Nizento DS?
- No. Every client chooses: the system online, or on a Nizento DS in your own office, where it carries on when the internet goes down.
- Does the AI decide anything?
- No. Money is never approved by the machine alone. A person approves each step, and every step records who did it and when.
- What happens after launch?
- You have a named person, an alert when anything stops, and a monthly review.
We hold no certifications and claim none. “Data Secure” is the product’s name, not a guarantee.
Start with one conversation
Build what’s next for your organisation.
From enterprise software to AI, let’s create technology around the way you work.