Most operations questions come down to the same worry: will adopting a system mean a painful switchover before anything gets better? The honest answer is that the change is a sequence, not a switch. To make that concrete, it helps to fix three reference points and compare them directly.
- Now: before Jiri
- Day 0: first day on Jiri
- Day 90: once the habit holds
Here is the short version, and the rest of the post is the detail behind it. Now is paper, spreadsheets, and status that has to be chased. Day 0 is the same people doing the same work, with every task captured once, given an owner and a due date, and rolled up so status is read rather than chased, and it needs no new hardware. Day 90 is when the captured record lets the system assemble reports, roll summaries up on a schedule, chase quiet work, and, for physical operations, read counts from a device instead of a person.
What does a factory order look like at each stage?
For a manufacturing order, the journey runs from taking the order to closing it out. The table below follows that one order through the backend. Notice that the jump from now to day 0 uses no sensors at all: every fact that lived on a phone call or a register becomes an assigned, timed, recorded task. The move to day 90 is then field by field, as scans and signals take over the counting a person used to do.
| Step | Now | Day 0 with Jiri | Day 90 |
|---|---|---|---|
| Take the order | Phoned in, jotted in a register. | Entered once, on the owner's board instantly. | Buyer can submit it through a form. |
| Check stock and capacity | Phone the godown, count by hand. | Capture tasks for stock and machine time, with due dates. | Read from scans and a machine signal, no person. |
| Quote the buyer | Price and date sent on a guess. | Price set by a person, quote recorded with a real date. | Same, and the system chases the buyer to reply. |
| Run production | Progress scribbled on a sheet, often late. | Recurring "units made?" task each shift, late shifts flagged. | Counts stream from the machine. |
| Material runs short | Phone chain: manager, owner, supplier, chase. | Owner approves, the approval is logged, reminders chase the confirm. | System messages the supplier on approval. |
| Dispatch the goods | Written in a gate register, handover on trust. | Marked dispatched with the vehicle, recorded. | Driver and manager scan to confirm both ends. |
| Track and close | Buyer phones to ask, sales phones around. | Buyer page shows status, delivered closes it. | Live status appears automatically. |
The pattern is that Jiri Ops is process-paced: a step moves when a task is done, a scan happens, or a device reports, not when someone remembers to update a sheet. The big gain on day 0 is simply that every off-system fact is now an assigned, timed, recorded task.
What does an NGO programme look like at each stage?
For programme and back-office work the journey is human-paced rather than process-paced, so day 90 does not mean machines replace the logging. It means the system assembles, rolls up, reminds, and surfaces what people have already logged. The work, the field visits, and the reviews stay human throughout.
| Step | Now | Day 0 with Jiri | Day 90 |
|---|---|---|---|
| Create the work | A chat or a notebook line, only two people know. | A small form, now a real item on the org tree. | Same form, with role-based templates. |
| Assign and accept | Handed over verbally, easy to forget. | Downward goes live, upward needs acceptance, creator is notified. | Same, and quiet pending asks get nudged. |
| Do the work, log activity | Notes in a notebook, photos on a phone. | Place, count, note, photos logged to the programme trail. | Same capture, the habit that powers the rest. |
| Submit and review | Mentioned in passing, maybe not checked. | Owner submits, assigner confirms or returns with a note. | Same, and waiting reviews surface on the board. |
| See who is doing what | Phone calls and meetings, no live picture. | Each view rolls up from the record, late items flagged. | A bottleneck lens shows what is stuck. |
| Report up | Assembled by hand, re-keyed at each level, late. | Built from the trail, a person sends it. | Summaries roll up on a schedule. |
| Keep the record through change | A leaver takes files and knowledge with them. | History stays on the record, views re-scope on their own. | Same, now deep enough for trends and audits. |
Why is day 0 the same hardware but a different operation?
Day 0 changes nothing about who does the work or what tools they touch in the field, and it adds no sensors. What it changes is where the record lives. Today the record of what is happening is scattered across a notebook, a chat thread, a drawer, and one person's memory, so nobody sees the whole picture without chasing it, and the chasing is itself work. On day 0 every piece of that work becomes one tracked item with an owner and a due date, and status is read from that record instead of gathered by phone. For a team that has run on paper for years, that shift alone is the difference between guessing and knowing.
Why does day 90 arrive field by field instead of all at once?
Day 90 builds on the capture habit rather than replacing it, which is why it is incremental. Once work is reliably captured, the system can assemble reports from the record, roll summaries up on a schedule, and chase work that has gone quiet. For physical operations it can go further: a manual count a person did at the start, reading stock or confirming a handover, can be handed to a scan or a device signal, while the flow underneath stays identical. A step that was a tracked human task simply becomes a tracked automatic one. Nothing forces every step to automate, and nothing automates before its capture is trusted.
What stays human across all three stages?
The judgement stays human at every stage. Automation here is not a goal for its own sake; it removes the repetitive relay work of asking, copying, calculating, and chasing, so people are left with the parts that genuinely need a decision. Pricing a quote, approving a purchase, deciding what to do when something goes wrong, doing the field work, and reviewing submitted work all stay with people in every version. A good operations tool makes those decisions faster to reach and keeps a clean record of them. It does not try to make them for you. To be clear about what powers the day 90 automation: Jiri uses rules-based automation today, not AI, and AI or voice is a planned future direction rather than a current feature.
Where should a team start?
Start by making the work visible, not by buying equipment. Pick the handoffs that hurt most, the places where things get lost or run late, and make those a tracked task with an owner and a due date. Let the live view replace the status phone calls. Only once that habit holds is it worth asking which manual steps are worth automating, and in what order. The sequence is the whole point: see the work, then assign it, then automate the parts that earn it. You start exactly where you are.
Common questions
What do now, day 0, and day 90 mean for Jiri?
Now is how a team works before Jiri: paper, spreadsheets, phone calls, and status that has to be chased. Day 0 is the first day on Jiri, when the same people do the same work but every task and field activity is captured once, assigned with a due date, and rolled up the org tree, with no new hardware. Day 90 is once the capture habit holds, when the system assembles reports, rolls summaries up on a schedule, chases quiet work, and, for physical operations, lets device signals take over manual counting.
Does Jiri need new hardware or sensors on day 0?
No. Day 0 value does not depend on sensors, scanners, or integrations. The only change on the first day is that each piece of work becomes a tracked task with a named owner and a due date. Device signals are a day 90 option for physical operations, not a requirement to start.
What stays human after day 90?
The judgement. Pricing, approvals, deciding what to do when something goes wrong, the field work, and reviews all stay human. Jiri removes the relay work of asking, copying, calculating, and chasing, so people are left with the parts that need a decision.