The founder’s story
Hashboard is the operations board for dangerous-goods freight. It was built by someone who does the work: Hafiyyan, an operations lead at a freight forwarder in Singapore, where a team of two moves gases and chemicals for customers around the world.
The first question
Every shipment carries a long tail of work: the booking, the dangerous-goods approvals, the documents, the cut-offs and the follow-up actions that keep it moving. For a small operations team that tail used to live in mailboxes, spreadsheets and memory.
On 11 August 2026, at 10:28 in the morning, Hafiyyan typed a question to an AI assistant:
“How do I create a ui to be able to track all my shipments and see the status of the shipments as they progress through the process”
A “ui” is a screen. About twenty minutes later he asked how to run a working version, and a minute after that he wrote: “i tried clicking but nothing works”. He was not a programmer, and he did not pretend to be. He knew exactly what the screen had to show.
The next morning the first working screen was saved to a code repository. It was simple: one flat row for each shipment. On 14 August it was on the web, and within days it had a sign-in of its own.
The name
The name is his own. In his words:
“I initially started with Hashboard as H is also the beginning of my name Hafiyyan, and I saw that my manager used the S# behind every shipment, So i derived the word ‘Hashtag” and created the dashboard as ‘Hashboard’.”
Every shipment already carried its own number, and a hash sign is how people tag a thing so it can be found. The board that holds every tagged shipment became the Hashboard.
How it is built
Hafiyyan is the product owner and the logistics expert, and every business rule on the board is his decision or was approved by him. AI agents write the software and a different agent checks it before anything reaches the live board. His reason for mixing them: “both have to work together as i feel the different agents can produce better results when working together.”
The same care applies to what the board reads. Mail is read into proposed changes, and a person confirms each one. His reason: “AI tends to make mistakes when doing things such as reading emails or crafting emails.” Nothing is sent to anyone without a person approving it.
What it does today
Every shipment has one record, from booking to delivery: containers, cargo, documents, approvals and cut-offs, all in one place. Every date says whether it is planned, estimated or actual, so nobody mistakes a guess for a fact, and every change names the person who made it.
The customer’s daily report is written from the board each afternoon, and the weekly movement plan comes from the same records, so a shipment corrected once is corrected everywhere it appears. Carrier and haulier mail is read for the facts that matter, and the letters that must be ready before a ship berths are prepared in advance. It works on a desk and on a phone, and an assistant in a chat app answers questions about the shipments, but only when asked.
It is built for the way Singapore freight already works: PSA berthing, trucking hours, haulier instructions and the carriers’ dangerous-goods forms.
First verdicts
On 13 October 2026 two verdicts came from outside the build. Linde, the world’s largest industrial gas company, praised the daily shipment tracker it receives. That file comes wholly from Hashboard: the board’s own rows written into Linde’s own workbook, nothing typed by hand.
The same day his manager, who had wanted a tool that answers “where is the shipment” from a phone, said he was impressed by how the chat assistant had developed: two days after it went live it had shown him a shipment’s details and built a carrier’s dangerous-goods declaration on command. He asked for one thing more, a status line that says more than the team’s own notes, and that became the next change.
Who it is for
Hashboard is for the people who have to know where every box is: operations teams at forwarders who handle dangerous goods, where a missed cut-off or an out-of-date arrival date costs real money and trust. It is for a small team that wants the record, the dates and the reports to agree with each other without retyping them.
What Hafiyyan brought was the work itself: which document is final, which date wins when two sources disagree, what the customer must see at 17:00 each day, and who may press send. A person who knows the work built the tool the work needed.
← Back to the landing page