Blog
Running the books, and training the models
Practical writing on accounting, VAT and payroll in Bangladesh, and on building the datasets and fine-tuned models that make AI useful.
Will AI Make You Jobless? The Honest Answer Is Yes and No
AI will eliminate some jobs, change most of them, and create new ones. What the data from the WEF and IMF actually says, which kinds of work are most exposed, and what individuals and businesses can do about it.
Read article →
Raw Data Sharing vs Tool Calling: Two Ways to Give AI Your Business Data
You can give an AI your data by pasting or uploading it, or by letting it call tools that fetch exactly what it needs. The difference decides whether your AI leaks data, makes up numbers, or can be trusted with real questions.
Read article →
Fine-Tuning vs RAG vs Prompting: How to Decide Before You Spend Money
Fine-tuning gets proposed for problems retrieval would solve in a week, and retrieval gets bolted onto problems only training can fix. A practical way to tell which one your use case needs.
Read article →
What Makes a Good Fine-Tuning Dataset (and Why Most Aren't)
A fine-tuned model learns exactly what its examples teach, mistakes included. Here is what separates a dataset that produces a reliable model from one that produces an expensive disappointment.
Read article →
How to Choose Accounting Software in Bangladesh: 7 Questions to Ask Before You Buy
Most accounting software demos look the same. These seven questions separate the systems built for how Bangladeshi businesses actually file VAT, withhold tax and run payroll from the ones that just translate the menus.
Read article →
The Real Risks of Doing AI Without Professionals (What the Demo Doesn't Show You)
Plugging a chatbot into your business takes an afternoon. Keeping it from leaking data, inventing policies, or deleting things takes design. Here are the risks that don't show up in a demo, and what professional handling actually changes.
Read article →
9 Real AI Failures and What They Teach Every Business
A chatbot that invented a refund policy, an agent that deleted a production database, a consultancy that refunded a government over fabricated citations. Nine documented AI failures, what went wrong in each, and the safeguard that would have prevented it.
Read article →
Can AI Really 10x Your Business? Where the Multiples Actually Come From
5x and 10x gains from AI are real, but they happen on specific workflows, not whole businesses overnight. Here's where the big multiples come from, where they don't, and how to stack them into growth that shows up on the P&L.
Read article →
Where to Draw the Line: Setting Boundaries for AI in Your Business
How much should an AI be allowed to see, say and do? A practical framework for setting boundaries on chatbots and AI agents, from read-only answers to actions that need a human's approval.
Read article →
How Businesses Lose Revenue to Poor Transparency — and Don't Even Know It
Most businesses don't bleed money in dramatic ways. They lose it quietly — in receivables nobody chased, discounts nobody claimed, and decisions made on numbers nobody trusted. A look at where the money actually goes.
Read article →
The 8th of the Month: What a Week-Long Month-End Close Really Costs You
If your books close a week late every month, you are not just behind on paperwork — you are making every big decision of the month on numbers nobody has checked. Here is where the week actually goes, and what it costs.
Read article →
The Next Generation of Business Will Be Agentic — What That Actually Means
Every business wave gets a name. This one is 'agentic AI' — software that doesn't just show you information but answers questions, does the looking-up, and stays inside rules you set. Here is what it changes in practice, and what it doesn't.
Read article →
Why AI Transformation Fails — and How to Get It Right
Most AI initiatives stall at the pilot stage. The difference between demos that impress and systems that deliver comes down to three decisions made before any model is built.
Read article →
AI Quality Control in Garment Factories: A Practical Guide
Manual inspection can't scale with buyer expectations. Here's how AI-powered QC actually works on a garment line — what it catches, what it costs, and how to run a pilot that proves ROI.
Read article →