Topic hub · Artificial intelligence
Artificial intelligence for business in Nepal: value, finance, risk and policy
Artificial intelligence has moved from the technology page to the board agenda. This hub gathers our analysis for owners, directors and finance leaders in Nepal: where the value sits in a mid-sized company, what changes in the finance function, how the tax office and lenders will use data, which controls are needed, and what national policy means in practice.
Start with the board questions
Begin with the flagship piece on the questions a board should ask, then the articles on value, build or buy, readiness and why programmes fail.
Finance, tax and lending
The finance series covers the CFO's office, the close, forecasting and controls. The tax and credit articles explain how the Inland Revenue Department and lenders read data, and what to do before they do.
Risk, sectors and policy
Governance and assurance come next, then a sector-by-sector view and what the National AI Policy 2082, Nepal Rastra Bank's guidelines and the budget mean for business.
Guides for business owners
Plain language: what it means, what to do and what to prepare.
- For business owners · 11 Nov 2026Why AI Programmes Fail, and What the Survivors Do DifferentlyMost AI programmes fail for management reasons, not technical ones. Ten failure patterns, how each shows up in a Nepali company, and the seven habits that separate the small group that gets results.
- For business owners · 11 Nov 2026Where the Value Is: Sizing AI's Real Payback in a Mid-Sized Nepali CompanyIn a low-wage, high-interest economy the payback from AI is rarely in headcount. It is in working capital, leakage, lost sales and avoided penalties. A method for sizing it in rupees, with a worked example for an NPR 80 crore distributor.
- For business owners · 11 Nov 2026Nepal's AI Opportunity: Delivery Base, Policy and the Jobs QuestionNepal is among the countries least exposed to AI and least prepared for it. That is both protection and risk. An assessment of the real opportunity in export services, the claims made for hydropower-fed compute, what the National AI Policy does and does not do, the evidence on jobs and migration, and a practical agenda for government, companies and individuals.
- For business owners · 11 Nov 2026The Data-Driven Tax Authority: How Analytics Changes Assessment Risk in NepalTax administrations worldwide now select cases by data matching and risk scoring, and Nepal's Inland Revenue Department has announced an AI-driven e-assessment system. What the tax office can already see, how selection is likely to change, and how a company should check its own data before the system does.
- For business owners · 11 Nov 2026Data Before Algorithms: Why Most Nepali Companies Are Not Ready for AI, and What Readiness CostsAI fails on bad data long before it fails on bad models. A five-layer readiness test for a Nepali company, the specific problems found in local books, what it costs to fix them, and the order to fix them in.
- For business owners · 11 Nov 2026Who Reviews the Machine? Controls Over AI-Generated NumbersWhen software prepares the reconciliation, drafts the report and proposes the journal, the old control of one person preparing and another reviewing no longer works as designed. A control framework for AI-assisted finance: what can go wrong, which controls address it, and what evidence a board, auditor or tax officer will expect.
- For business owners · 11 Nov 2026From Monthly Accounts to Continuous Insight: Redesigning the CloseMost Nepali companies see their monthly results three to six weeks late, or only at year end. The close is slow because of how the work is organised, not because of the software. How to redesign it step by step, where AI helps, and what a continuous close really means.
- For business owners · 11 Nov 2026Continuous Assurance: Testing Every Transaction Instead of a SampleSampling was a response to the cost of checking by hand. When every transaction sits in a database, a company can test all of them, every month, with rules a spreadsheet can run. What full-population testing finds, the twenty tests to start with, how to handle exceptions and where AI helps and misleads.
- For business owners · 11 Nov 2026The CFO's Office in 2030: What AI Changes in a Nepali Finance Team, and What It Does NotBy 2030 the routine half of a Nepali finance team's work will be done largely by software. The half that remains is judgement, control, explanation and accountability. A task-by-task view of what changes, what does not, and how to redesign the team.
- For business owners · 11 Nov 2026Can AI Be Trusted on Nepali Tax Law? A Test Protocol and an Answer KeyGeneral AI tools are structurally weak on Nepali tax law: little training material, annual rate changes, Nepali-language sources and a different calendar. Instead of asking you to trust anyone's test results, this article gives a twenty-question test with an answer key, a scoring method and rules for safe use.
- For business owners · 11 Nov 2026Build, Buy or Wait: A Decision Rule for AI Investment in a Small MarketMost Nepali companies should use what is already inside their software, buy a little, configure a little and build almost nothing. A five-option ladder, four tests and a contract checklist for deciding where each AI use case belongs.
- For business owners · 11 Nov 2026AI Risk Is a Governance Question: Accuracy, Confidentiality and Dependence at Board LevelAI is already in use in most Nepali companies, whether or not the board has approved it. The risks fall into three groups: wrong output, leaked information and dependence on vendors and on machines. A board-level framework, a one-page policy, the legal position in Nepal and the questions directors should ask.
- For business owners · 11 Nov 2026The AI Question Every Nepali Board Should Be AskingMost boards ask whether the company should use AI. The better question is which decisions will be made differently, on what data, and who answers for the result. Eight questions for directors in Nepal, with what a good answer looks like.
- For business owners · 11 Nov 2026Forecasting Under Uncertainty: Where Machine Prediction Beats Judgement, and Where It Does NotStatistical and machine forecasts beat human judgement on stable, repeated, data-rich questions, and fail at structural breaks, which Nepal has often. A practical guide to what to forecast by machine, what by judgement, how to combine them, and how to measure whether a forecast is any good.
- For business owners · 11 Nov 2026How Lenders Will Read Your Business: AI in Credit Decisions and What Borrowers Should Do NowLending in Nepal has rested on collateral and relationships. Credit scores, account data, tax filings and digital payment trails are moving it toward data. What a lender's model can see about your business, what it rewards and penalises, the rights the central bank's AI guideline gives borrowers, and a twelve-month plan to become a borrower the data favours.
- For business owners · 11 Nov 2026AI Across Nepal's Sectors: Where It Matters FirstAI will not arrive evenly across Nepal's economy. It reaches first the sectors with digital data, scale, competitive pressure and people who can use it. A sector-by-sector assessment for twelve parts of the economy: what is realistic now, what blocks it, the first use worth trying and what to leave for later.
How we can help
- Data analytics and reportingClean data, management reporting and dashboards
- Process automationRemoving manual work from finance processes
- Virtual CFOFinance leadership, forecasting and board reporting
- Internal auditControls, testing and assurance over processes
Tools and references
- Business health checkA first read on finance and controls
- Cash flow calculatorA baseline before any forecast model