Building Bankmore and How AI Finally Solves Small Business Accounting

Last Tuesday, I watched my friend Sarah spend three hours trying to reconcile a single month’s expenses in FreeAgent. She’s a brilliant product designer who runs a thriving consulting practice, but there she was, squinting at bank statements and frantically searching through Gmail for receipt attachments from two months ago. When she finally found that elusive £47 software subscription buried in her spam folder, she looked at me and said, “There has to be a better way.”

She’s right. And that better way is exactly what we’re building with Bankmore.

The thing is, Sarah’s experience isn’t unique. Over the past decade of running Devular, my technology consultancy, I’ve watched dozens of smart entrepreneurs get completely derailed by basic accounting tasks. These are people who can design beautiful user interfaces, architect complex software systems, or negotiate million-dollar deals, yet they’ll spend six hours building a feature that saves their customers thirty seconds, then lose entire afternoons hunting for receipts because their accounting software can’t talk to their email.

It’s not their fault. Traditional accounting software was designed for accountants, not entrepreneurs. Having worked with large UK media companies on big data projects, I’ve seen how enterprise financial systems assume dedicated bookkeeping staff, established workflows, and the patience to learn byzantine categorization systems. But most small business owners are wearing fifteen different hats, working from coffee shops, and just trying to keep their businesses running while staying compliant with tax regulations.

That disconnect between how accounting software works and how entrepreneurs actually work is exactly what Bankmore is designed to solve.

The Hidden Complexity That’s Killing Productivity

Here’s what most people don’t realize about small business accounting: the software isn’t the hard part. Xero and FreeAgent are actually quite powerful once you understand them. The hard part is the endless context switching, data hunting, and manual reconciliation that happens around the software.

Let me walk you through Sarah’s typical monthly routine before Bankmore. She’d export bank statements, download transaction CSVs, then spend hours matching payments to invoices. But that’s just the beginning. For every transaction, she’d need to:

Find the receipt: Usually buried somewhere in Gmail, often forwarded from an assistant or team member who made the purchase. Sometimes it’s a PDF attachment, sometimes it’s an image, sometimes it’s just text in the email body.

Determine the category: Is this software expense “Computer Software” or “Professional Services”? What about that conference ticket, is it “Training” or “Marketing”? Every accountant has slightly different preferences, and getting it wrong means corrections later.

Split complex transactions: That £247 Amazon order contains office supplies, a book (training expense), and some equipment. Good luck figuring out the breakdown three weeks later when you’ve forgotten what you even ordered.

Handle international payments: Currency conversions, foreign transaction fees, VAT implications for EU purchases. Each one requires research and often a call to her accountant.

Track project expenses: Which client should be billed for that research tool subscription? Was that travel expense for the London project or the Manchester one? Without meticulous note-taking in the moment, this information is lost forever.

The result? Sarah typically spent 4-6 hours per month on accounting tasks, usually on weekends when she could focus without interruption. But more importantly, the cognitive overhead was constant. Every purchase required mental bookkeeping: “I need to remember to save this receipt, categorize this correctly, and bill this to the right project.” It’s the kind of background mental load that exhausts entrepreneurs and prevents them from focusing on what actually grows their business.

How AI Changes the Game Completely

This is where Bankmore’s approach gets interesting. Instead of trying to make traditional accounting software slightly easier to use, we’re using AI to eliminate most of the manual work entirely.

Here’s how it actually works in practice. When Sarah makes a purchase now, Bankmore immediately scans her email for receipts, extracts the relevant information, and suggests categorizations based on her historical patterns and business context. But it goes deeper than simple automation.

Smart receipt hunting: The AI doesn’t just look for emails with “receipt” in the subject line. It understands that confirmation emails from SaaS tools are functionally receipts, that payment confirmations from PayPal contain expense information, and that calendar invites for paid events often include pricing details. It can correlate a Stripe payment notification with a software subscription renewal that happened three days earlier.

Context-aware categorization: Instead of making you choose between “Computer Software” and “Professional Services,” Bankmore learns your specific business patterns. It knows that payments to Figma are design tools (professional services for Sarah’s design work), while payments to Linear are project management software (business overhead). It remembers that conference tickets in London are usually marketing expenses, while ones in San Francisco are typically training.

Intelligent expense splitting: When Sarah orders £300 worth of supplies from Amazon, Bankmore can analyze the order confirmation email, identify individual items, and suggest how to split the expense across categories and projects. It knows that ergonomic keyboards are equipment, business books are training, and sticky notes are office supplies.

Automated client billing: Perhaps most importantly, Bankmore tracks which expenses should be billed to specific clients. It learns from Sarah’s patterns, travel expenses during project weeks typically get billed to the client she was visiting, software purchases for specific deliverables get categorized as reimbursable project costs.

The result? Sarah’s monthly accounting routine went from 6 hours to about 20 minutes of review and approval. But the bigger win is cognitive: she no longer thinks about accounting when making purchases. She just buys what she needs and trusts that Bankmore will handle the administrative overhead.

The Magic Happens in the Details

What makes Bankmore genuinely useful isn’t the big AI features, it’s the hundreds of small details that eliminate friction from everyday accounting tasks.

Email integration that actually works: We don’t just scan for receipts; we understand email threading, can read forwarded messages, and know how to extract pricing from booking confirmations, invoice emails, and payment notifications. When Sarah’s assistant forwards a receipt, Bankmore knows to attribute the expense to Sarah’s account, not her assistant’s.

Natural language interaction: Instead of forcing users to learn accounting categories, Bankmore lets them ask questions in plain English. “How much did I spend on software last quarter?” “Which client expenses haven’t been billed yet?” “What’s my biggest marketing expense?” The AI translates these questions into proper financial queries and presents results in understandable language.

Proactive accountant communication: This is the feature that consistently surprises users. Bankmore doesn’t just organize your expenses, it actively communicates with your accountant about unusual transactions, deadline reminders, and potential deductions you might have missed. It’s like having a bookkeeper who never sleeps and never forgets to follow up.

Dynamic reporting: Rather than static monthly reports, Bankmore generates interactive dashboards that let you drill down into specific expense categories, time periods, or projects. The AI creates bespoke visualizations based on what’s most relevant to your business, cash flow trends for service businesses, project profitability for consultants, expense efficiency for e-commerce operations.

The Broader Implications for Small Business

But here’s what I find most interesting about building Bankmore: it’s revealing how much administrative overhead has been holding small businesses back. When accounting tasks take 20 minutes instead of 6 hours, entrepreneurs don’t just get their evenings back, they start making different business decisions.

Sarah now tracks expenses in real-time instead of monthly, which means she caught a duplicate software subscription that had been charging her for eight months. She identifies profitable projects faster and can price future work more accurately. She experiments with new tools and services because the administrative cost of evaluating them is essentially zero.

Multiple Bankmore users have told us they’re taking on larger projects because they’re confident they can track complex, multi-vendor expenses without drowning in paperwork. Others are expanding internationally because they’re no longer afraid of foreign transaction complexity.

The AI isn’t just making existing processes faster, it’s enabling business strategies that were previously too administratively complex to pursue.

Building for the Reality of Modern Business

One thing that became clear early in Bankmore’s development is that traditional accounting software assumes a business model that barely exists anymore. It assumes clear boundaries between business and personal expenses, single-currency transactions, predictable vendor relationships, and employees who understand accounting categories.

Modern entrepreneurs don’t work that way. They buy software on personal credit cards and reimburse themselves later. They attend conferences that are part networking, part training, part marketing. They work from co-working spaces in three different countries and pay for services in six different currencies. They collaborate with freelancers, agencies, and contractors whose invoicing systems range from professional to chaotic.

From my experience building cross-platform apps and working with distributed teams, I’ve seen how the traditional boundaries between work and life, business and personal, local and international have completely blurred. Bankmore is designed for this reality. It handles mixed personal/business transactions, understands that the same expense might serve multiple business purposes, and can work with whatever financial infrastructure an entrepreneur has cobbled together.

Multi-account intelligence: Bankmore connects to all your accounts, business checking, personal credit cards, PayPal, Stripe, cash transactions recorded via photos, and maintains a coherent view of business expenses regardless of payment method.

Collaborative expense handling: When team members make purchases, Bankmore can process their receipts, categorize expenses according to your business rules, and automatically request reimbursement through your preferred system.

International business support: The AI understands VAT implications, currency conversion timing, and country-specific expense categories. It knows that meals in London are 100% deductible while meals in San Francisco are only 50% deductible, and it tracks these nuances automatically.

The Technical Architecture That Makes It Work

Building an AI system that actually understands business finances required solving some interesting technical challenges. The AI needs to be smart enough to understand context and patterns, but reliable enough to handle financial data that entrepreneurs depend on for tax compliance.

We use a combination of large language models for natural language processing and specialized financial AI models for transaction categorization. But the real magic happens in the integration layer that connects emails, bank feeds, accounting software, and business intelligence tools into a coherent system.

Real-time data processing: Instead of batch processing monthly exports, Bankmore processes transactions as they happen. When your card is charged, the AI immediately starts hunting for receipts, categorizing the expense, and updating your financial dashboards.

Context preservation: The AI maintains a persistent understanding of your business, your clients, projects, vendors, expense patterns, and accounting preferences. It learns from every transaction and gets smarter about categorization and billing recommendations over time.

Privacy-first architecture: Financial data never leaves secure, encrypted environments. The AI processes your information to provide insights and automation, but we’ve designed the system so that sensitive financial details are never stored in readable form or shared with external services.

What We’re Learning from Early Users

The most surprising feedback from Bankmore’s beta users isn’t about specific features, it’s about how the app changes their relationship with business finances entirely.

Before Bankmore, most entrepreneurs treated accounting as a necessary evil, something to endure and minimize. Now they’re actively curious about their financial patterns. They’re asking questions like “Which types of marketing spend actually generate clients?” and “How does my expense efficiency compare to industry benchmarks?”

One user told us that Bankmore made her realize she was spending £400 per month on software subscriptions she’d forgotten about. Another discovered that his most profitable projects consistently involved specific types of travel expenses, leading him to adjust his service offerings. A third user found that she was dramatically undercharging for projects that required specialized software licenses.

The AI isn’t just processing expenses, it’s helping entrepreneurs understand their businesses in ways that weren’t practical when accounting was a manual, monthly chore.

The Road Ahead for Expanding Beyond Basic Accounting

Currently, Bankmore focuses on expense management and basic financial reporting, but we’re seeing opportunities to expand into broader business intelligence and financial planning.

Predictive cash flow: Using historical patterns and current project pipelines, the AI can forecast cash flow needs and suggest optimal timing for major expenses or investments.

Automated tax optimization: Instead of waiting until tax season to identify deductions, Bankmore can suggest tax-efficient expense timing and identify overlooked deductible activities throughout the year.

Business performance insights: By analyzing expense patterns alongside revenue data, the AI can identify which business activities generate the highest returns and suggest resource allocation adjustments.

Integrated financial planning: Rather than treating accounting as record-keeping, Bankmore aims to become a forward-looking financial advisor that helps entrepreneurs make better business decisions based on real financial data.

Why This Matters Beyond Accounting

What we’re really building with Bankmore isn’t just better accounting software, it’s infrastructure that lets entrepreneurs focus on entrepreneurship instead of administration. When basic business operations are automated and intelligent, entrepreneurs can experiment more, take bigger risks, and pursue opportunities that would have been administratively prohibitive.

I think about Sarah, who’s now taking on international clients because she’s not afraid of complex expense tracking. Or the consultant who’s testing new service offerings because he can accurately track profitability in real-time. Or the designer who’s finally tracking time properly because the administrative overhead of detailed expense tracking has disappeared.

Better accounting software might save entrepreneurs a few hours per month. But intelligent business infrastructure that thinks alongside entrepreneurs and handles administrative complexity automatically? That changes what kinds of businesses are possible to build and run.

The goal isn’t to replace accountants or eliminate financial oversight, it’s to give entrepreneurs the same kind of intelligent business infrastructure that large companies take for granted, but tailored for the reality of how small businesses actually operate.

That’s the future we’re building with Bankmore. Not just better accounting, but fundamentally more intelligent business operations that scale with entrepreneurs as they grow.