Actuality

Actuality: Protecting Business Owners Through AI-Powered Inventory Intelligence

1. What Inspired Me

My inspiration for building Actuality came from hearing stories about business owners in Nigeria who struggle with inventory losses, suspected employee theft, stock manipulation, and other fraudulent activities within their businesses.

Many entrepreneurs invest their savings, time, and energy into building their businesses. They purchase goods, employ people they trust to manage their stores, and depend on those employees to keep accurate records of sales and stock movements. However, when business owners cannot continuously monitor their inventory, they may find it difficult to know exactly what is happening in their absence.

A shop owner might purchase 100 units of a product but later discover that only 75 units remain, even though the recorded sales do not explain the difference. In some cases, the discrepancy might result from damaged goods, recording mistakes, unrecorded sales, or theft. Without reliable records, identifying the actual cause can be difficult.

What concerns me most is that business owners may not discover these discrepancies until they have accumulated into significant financial losses. For small businesses operating with limited capital and narrow profit margins, these losses can affect their ability to restock, pay employees, expand, or even remain in business.

A February 2026 report by the Centre for the Promotion of Private Enterprise, covered by TheCable, highlighted employee-related fraud as a serious challenge for Nigerian MSMEs, including inventory theft, diversion of sales proceeds, and falsification of financial records. This reinforced the importance of addressing the problem through stronger accountability and better business controls. https://www.thecable.ng/cppe-msmes-lose-n5trn-annually-to-occupational-fraud-employee-corruption/

I began to see inventory management as more than a way to count products. I saw an opportunity to help business owners protect their investments, improve transparency, and make informed decisions using technology.

That is the idea behind Actuality.

2. The Problem I Want to Solve

Many small and medium-sized businesses still depend on handwritten records, spreadsheets, or inventory systems that do not provide sufficient visibility into daily operations. Some business owners must rely heavily on the information provided by their stockkeepers because they cannot personally supervise every transaction.

This creates several challenges:

• Inventory theft and unexplained stock shortages: Goods may go missing without a clear record of when or how the discrepancy occurred.
• Record manipulation: Inadequate transaction controls can make it difficult to distinguish genuine errors from deliberate changes to inventory records.
• Unrecorded or falsified sales: Sales may not be properly reflected in inventory records, making reconciliation difficult.
• Unauthorised stock adjustments: Products may be removed, written off, or adjusted without appropriate approval.
• Delayed detection: Business owners may discover discrepancies only during periodic stocktaking, after losses have accumulated.
• Limited remote oversight: Owners who manage multiple shops or cannot remain on-site throughout the day may lack timely information about their stock.

The challenge is not that every stockkeeper is dishonest. Many employees are hardworking and trustworthy. The problem is that systems based solely on trust, manual records, and occasional supervision can leave both honest employees and business owners vulnerable to mistakes, disputes, and abuse.

I want Actuality to help close that gap by making inventory activities more transparent, traceable, and accountable.

3. My Solution: Actuality

Actuality is an AI-powered, offline-first inventory management system designed to help business owners maintain control over their stock and identify suspicious discrepancies earlier.

Its central idea is simple: every important inventory movement should have a clear record, an identifiable source, and an appropriate level of accountability.

Rather than simply showing how many products remain in a store, Actuality is intended to help owners understand how inventory changes over time and whether recorded transactions reconcile with the physical stock.

The system is being developed around an accessible desktop application, with a focus on businesses that need practical inventory tools without depending on a continuous internet connection.

My planned approach includes the following capabilities:

1. Inventory movement tracking

Record stock received from suppliers, items sold, goods returned, damaged products, and authorised adjustments. These records can help owners understand why inventory balances change.

2. Staff accountability and access control

Introduce individual staff accounts, role-based permissions, and transaction records so that sensitive actions can be associated with the appropriate user. This would make it more difficult for unauthorised changes to go unnoticed.

3. Stock reconciliation

Compare expected stock balances against physical counts. When a discrepancy appears, Actuality could help the owner investigate the difference instead of relying solely on manual calculations.

4. Suspicious activity alerts

Use configurable rules and, as the system matures, AI-assisted anomaly detection to flag unusual patterns. Examples could include repeated stock adjustments, unusually high write-offs, frequent reversals, or discrepancies concentrated around particular products or periods.

These alerts would indicate activities requiring investigation, not automatically establish that an employee has committed fraud.

5. Business intelligence and reporting

Provide owners with understandable summaries of stock movement, inventory discrepancies, product performance, and recurring operational issues to support better decisions.

6. Offline-first operation

Design the core desktop workflows to function without a continuous internet connection, making the system more practical for businesses operating in environments where connectivity may be unreliable.

Together, these capabilities would help move inventory management away from a system of occasional checks and towards continuous, evidence-based oversight.

4. How I Built the Project

I approached Actuality as a practical technology solution to a real business problem rather than as an AI project looking for a problem to solve.

I chose to begin with an offline-first desktop application because accessibility and reliability are important considerations for small businesses. A solution that requires constant connectivity or complicated infrastructure may be difficult for some businesses to adopt.

The initial application uses Python, PyQt6, and SQLAlchemy. PyQt6 provides the desktop user interface, while SQLAlchemy supports structured database operations and data management.

My development approach has focused on building a usable foundation for inventory operations before expanding the intelligence layer.

The work involves several important stages:

1. Inventory data management: Establish a structured way to organise products and their stock balances.
2. Transaction tracking: Create a foundation for recording inventory movements and maintaining more consistent records.
3. User-friendly workflows: Make the system accessible to business owners and employees who may not have advanced technical skills.
4. Reporting and visibility: Present inventory information in a way that helps owners understand their operations.
5. AI and backend expansion: Plan for additional intelligence capabilities and a backend architecture that can support more advanced analytics as the product develops.

The longer-term technical direction includes AI-assisted inventory analysis, anomaly detection, and intelligent recommendations. These capabilities require suitable transaction data, reliable business rules, testing, and validation before they can be trusted in real business environments.

My goal is not merely to build software that stores inventory figures. It is to develop a system that helps businesses understand what is happening to their stock and gives owners stronger tools for accountability.

5. What I Have Learned

Working on Actuality has strengthened my understanding of the relationship between technology, business operations, and trust.

First, accurate data is the foundation of useful AI. An AI model cannot reliably identify suspicious inventory patterns if transactions are missing, stock balances are incorrect, or staff actions are not recorded properly. Before adding sophisticated algorithms, the system must establish a dependable record of business activity.

Second, fraud prevention requires more than artificial intelligence. Many inventory discrepancies can be addressed through straightforward controls: recording every stock movement, restricting sensitive operations, requiring approval for certain adjustments, and regularly reconciling physical inventory with recorded balances. AI can complement these controls, but it cannot replace them.

Third, technology must work in the environment where people need it. An advanced system that is too expensive, difficult to operate, or dependent on unreliable internet access may fail to solve the problem. Practicality and accessibility are therefore central to Actuality's design.

Fourth, transparency protects honest employees as well as business owners. Clear records can help identify genuine mistakes, document authorised actions, and prevent employees from being unfairly blamed for discrepancies they did not cause.

Finally, I have learned that successful innovation requires understanding the people who will use the product. Business owners need more than dashboards and reports; they need clear explanations, trustworthy records, and actionable information.

6. The Challenges I Face

Developing Actuality presents several technical and practical challenges.

Building reliable inventory records

One of the most important challenges is ensuring that inventory records accurately reflect real-world transactions. Goods may be damaged, returned, transferred, or sold through different channels. The system must account for these activities to reduce false discrepancies.

Distinguishing fraud from ordinary errors

A stock shortage does not automatically mean theft has occurred. Data-entry mistakes, supplier delivery errors, damaged goods, and delayed transaction entries can produce similar patterns.

Developing useful fraud-risk indicators therefore requires careful rules, contextual information, and testing. AI-generated alerts must be treated as leads for investigation rather than proof of wrongdoing.

Protecting transaction integrity

If someone with sufficient access can alter or delete transaction records without leaving evidence, the system's value as an accountability tool is weakened.

I therefore recognise the importance of designing appropriate access permissions, audit trails, approval workflows, backups, and protections against unauthorised modification.

Making the product affordable and accessible

Many small businesses have limited budgets and may not have dedicated IT personnel. Actuality must provide meaningful value without introducing unnecessary complexity or expensive infrastructure.

Its offline-first approach is intended to help address some connectivity challenges, although reliable backups, device security, and controlled synchronisation would also need consideration as the product evolves.

Developing and validating the AI component

Reliable anomaly detection requires suitable data representing ordinary inventory operations and the different types of discrepancies that businesses experience. It is also necessary to measure whether the system identifies useful patterns without generating too many false alarms.

This means the AI component must be developed progressively and validated with real operational feedback rather than relying on assumptions about how fraud occurs.

Building trust and encouraging adoption

Business owners must trust the system with important operational information. Employees must also understand how it works and how their activities are recorded.

I believe clear communication, appropriate privacy protections, transparent records, and practical demonstrations will be important in building that trust.

7. My Vision for the Future

My vision is for Actuality to become a trusted inventory intelligence platform for small and medium-sized businesses, beginning in Nigeria and eventually reaching other markets with similar operational challenges.

I want a business owner to be able to check stock levels, review inventory movements, investigate discrepancies, and understand unusual patterns without needing to be physically present at the store every minute of the day.

Over time, Actuality could evolve to include stronger audit trails, remote owner reporting, multi-store oversight, AI-assisted anomaly detection, demand forecasting, and intelligent replenishment recommendations.

I also want the platform to support a culture of accountability in which business owners can protect their investments without treating every employee as a suspect. The objective is to build a system in which honest activity is properly documented and questionable discrepancies are easier to investigate.

Ultimately, I hope Actuality will help entrepreneurs preserve their capital, improve operational discipline, and focus more of their energy on growing their businesses.

8. Why Actuality Matters to Me

Actuality represents my belief that technology should solve problems that matter to ordinary people and businesses.

For a small business owner, losing stock is not merely an accounting discrepancy. It can mean losing money needed to restock, support a family, pay salaries, or keep a business alive. When such losses remain hidden, the owner may struggle to understand why the business is not growing despite working hard.

I want to use my experience in software development, data analysis, and artificial intelligence to help address this problem with a solution that is practical, accessible, and built around accountability.

I understand that technology alone cannot eliminate fraud. However, well-designed systems can reduce opportunities for manipulation, improve the visibility of business activities, and help owners make decisions based on evidence rather than guesswork.

That is the purpose behind Actuality: to give business owners greater confidence in their inventory, stronger visibility into their operations, and better tools to protect the businesses they have worked hard to build.

Actuality — Truth. Now. Always.

Built With

• lama
• python
• tensorflow