How Businesses Can Use AI to Automate Operations, Reduce Costs, and Build Custom Software

Artificial intelligence is changing more than how people work. It is changing how businesses are built, how employees spend their time, how software is developed, and how companies compete.

For years, businesses have relied on employees to perform repetitive tasks, purchased software subscriptions to manage their operations, and adapted their workflows to fit the capabilities of existing systems. AI is beginning to challenge these assumptions.

A small business can now use AI to automate routine customer inquiries, assist with accounting, generate marketing content, analyze data, and even build software tailored to its specific needs. Tasks that previously required significant human effort or expensive software development may become considerably easier and cheaper to accomplish.

But the real opportunity is not simply using ChatGPT or adding AI to an existing system. It is rethinking how the entire business operates.

As a technology consultant who has worked in the software industry for years, I believe we are entering an era where businesses will need to ask a fundamental question: What would my business look like if AI existed?

This article explores how businesses can use AI to automate operations, reduce costs, build customized software, empower employees, and prepare for the future of work.

1. Is AI Just Another Technology Trend, or a Fundamental Shift in Business?

AI is not just another technology trend. I believe we are witnessing a fundamental shift in how businesses should operate.

To understand this change, consider the introduction of electricity.

When electricity was first introduced, people had concerns, fears, and misconceptions about what it would mean for society. Over time, electricity became an essential part of everyday life. Businesses no longer viewed it as an extraordinary technology. It became infrastructure.

I believe AI is following a similar path.

Today, many people associate AI with chatbots, content generation, or tools such as ChatGPT. However, these are only visible applications of a much broader technological shift.

AI is becoming an underlying capability that businesses can integrate into almost every department.

Consider the possibilities:

• Marketing teams can use AI to research customers, create content, and analyze campaigns.
• Accounting teams can use AI to assist with financial analysis and repetitive bookkeeping tasks.
• Customer support teams can automate basic inquiries.
• Software developers can use AI to accelerate software development.
• Business owners can use AI to analyze operations and identify opportunities for improvement.

The important distinction is that AI is not merely another feature businesses can purchase.

It is becoming a new way of producing work.

Businesses that understand this shift will begin redesigning their operations around AI capabilities. Those that continue operating exactly as they did before may find themselves struggling to compete with organizations that can deliver similar services faster and at a lower cost.

2. How Should a Business Start Automating With AI?

The first step is not purchasing an AI subscription. It is conducting an AI audit of the business.

Before implementing any technology, business owners need to understand how their businesses operate, what problems they face, and where time and money are being wasted.

From my experience in software development and business technology consulting, I believe businesses should begin by analyzing their existing processes.

The analysis should identify:

• Repetitive tasks that consume employees’ time.
• Bottlenecks that delay business operations.
• Tasks that require unnecessary manual intervention.
• Processes that are expensive to maintain.
• Activities that could be automated.
• Tasks that require human judgment.
• Workflows that could be redesigned entirely.

Once these bottlenecks have been identified, the next step is determining whether AI can handle the task.

For example, consider a business that receives hundreds of customer inquiries every month.

Employees may spend hours answering the same questions about product prices, business locations, operating hours, and delivery policies.

A business owner could identify these repetitive inquiries and introduce an AI-powered customer support system to handle them.

However, not every task should be automated.

A process involving a critical financial transaction, for example, requires a different level of oversight than drafting a marketing email.

The business should evaluate the task based on its complexity, risk, and need for human judgment.

A practical AI automation framework

Businesses can begin with four steps:

1. Analyze: Understand how the business currently operates.
2. Identify: Find repetitive tasks and operational bottlenecks.
3. Evaluate: Determine whether AI can perform the task safely and effectively.
4. Automate: Implement the solution, test it, and measure the results.

The most important lesson is to start with a problem, not a technology.

3. How Can Businesses Use AI to Save Time and Reduce Operating Costs?

AI creates an opportunity for businesses to reduce the amount of time employees spend on repetitive tasks.

Consider a small business with five employees. As the business grows, it may need additional employees to manage customer support, marketing, administration, and reporting.

Traditionally, the owner might hire additional staff to handle the increasing workload.

With AI, the business may be able to automate some of these activities instead.

Examples of business tasks AI can automate

Business function	Potential AI application
Soutien à la clientèle	Answer frequently asked questions and route complex inquiries
Marketing	Draft content, research audiences, and assist with campaigns
Comptabilité	Assist with transaction categorization and financial reporting
Administration	Draft emails, summarize documents, and organize information
Sales	Research prospects, qualify leads, and prepare follow-up messages
Human resources	Assist with job descriptions, onboarding documents, and internal inquiries
Rapports	Generate summaries and identify patterns in business data

The objective is not necessarily to eliminate employees.

Instead, businesses can reduce the amount of time spent on routine activities and allow existing employees to focus on higher-value work.

For example, a customer support employee who previously spent the entire day responding to basic inquiries could spend more time resolving complex customer problems and improving customer relationships.

AI can help businesses handle growing workloads without automatically increasing their operating costs at the same rate.

However, these savings are not guaranteed. Businesses still need to account for AI subscriptions, software development, integration, training, hosting, and human supervision.

The real savings come when AI is implemented effectively and the underlying processes are redesigned.

4. Can AI Help Businesses Reduce Their Software Subscription Costs?

One of the areas I believe AI will significantly disrupt is the traditional software subscription model.

Many businesses subscribe to numerous software applications to manage their operations.

They may have separate subscriptions for:

• Payroll.
• Point-of-sale systems.
• Customer relationship management.
• Inventory management.
• Comptabilité.
• Human resources.
• Project management.
• Marketing automation.

When business owners add up all these subscriptions, they may discover that software has become a significant recurring expense.

AI is creating a new opportunity: businesses can explore building customized software that combines some of these functions into an integrated solution.

From multiple subscriptions to customized software

Imagine a business paying for separate payroll, inventory, POS, and CRM applications.

Instead of continuing to use multiple systems, the business could assess whether an integrated application would better serve its needs.

With AI-assisted software development, a developer can use natural language descriptions, AI coding tools, and modern development frameworks to accelerate the creation of customized applications.

The business owner can describe the operational problems they want to solve, and the developer can translate those requirements into software.

The resulting system could combine several workflows into one application.

For businesses with the right requirements, this approach may reduce subscription costs and improve integration between departments.

The economics of custom software

Traditional software development can require substantial upfront investment.

AI-assisted development may reduce the time and effort needed to create certain applications, making customized solutions more accessible to smaller businesses.

Instead of paying recurring per-user subscription fees for every application, a business might pay for:

• Initial software development.
• Hosting infrastructure.
• Maintenance.
• Security updates.
• Technical support.

This does not mean custom software is always cheaper.

A subscription application may be more economical when the business needs a specialized, frequently updated service. Custom software also carries ongoing responsibilities for security, maintenance, backups, and compliance.

The key is to evaluate the total cost of ownership.

Businesses should not automatically assume that every software subscription needs to be replaced. They should identify which subscriptions solve important problems and whether a customized alternative would deliver better value.

5. Why Should Businesses Build Software Around Their Workflows Instead of Adapting to Generic Software?

Every business is unique in one way or another.

Two businesses operating in the same industry may have different processes, customer expectations, pricing structures, and operational requirements.

Yet many businesses are forced to adapt their operations to the limitations of generic software.

I believe this is an opportunity AI-assisted software development can address.

Rather than asking employees to change how they work to fit a software application, businesses can explore building systems that adapt to their specific workflows.

A practical example: The automotive dealership

Consider a car dealership.

In a traditional model, customers may need to visit the dealership to inspect available vehicles, ask questions, and determine whether a particular car meets their requirements.

Now imagine a dealership using an AI-powered platform that allows customers to:

• Search for vehicles based on their budget.
• Filter cars by model, year, fuel type, and other specifications.
• Identify vehicles that match their requirements.
• Explore available options before visiting the dealership.
• Make more informed purchasing decisions.

The customer can understand what is available before spending time traveling to the dealership.

This is an example of how technology can improve the customer experience, rather than merely digitizing an existing process.

The competitive advantage comes from understanding the customer’s problem and designing a better solution around it.

The future of customized business software

AI makes it increasingly possible for businesses to explore solutions that previously might have been too expensive to develop.

A construction company could have software designed around its project reporting and inspection workflows.

A school could have a system tailored to its enrollment, fee collection, and administration processes.

A logistics company could build an application around its delivery operations.

The possibilities are extensive.

However, customization should be driven by genuine business requirements. Building software for every minor preference can introduce unnecessary complexity.

The objective is to create a solution that improves the business, not simply to create more software.

6. Will AI Replace Employees, or Will It Change the Work They Do?

AI will automate some tasks that employees currently perform. However, I believe businesses should also consider how AI can help employees become more productive and valuable.

Many employees spend a significant portion of their working hours performing repetitive activities.

These may include data entry, basic customer inquiries, scheduling, report preparation, and document processing.

AI can potentially reduce the time required to complete these activities.

Employees can then focus on more important responsibilities.

For example:

• Customer support employees can focus on complex customer problems.
• Sales employees can spend more time building relationships and closing deals.
• Accountants can focus on financial analysis and decision-making.
• Marketing employees can focus on strategy and understanding customers.
• Managers can focus on improving business operations.

Why the human touch still matters

Even if a business automates basic customer support, customers still need to feel that the business understands them.

AI can answer simple questions such as:

• Where is your business located?
• What are your operating hours?
• How much does a product cost?
• What is your delivery policy?

However, when customers have complex concerns, complaints, or sensitive issues, human interaction becomes important.

The goal should be to combine the efficiency of AI with the empathy, judgment, and relationship-building abilities of people.

AI should help employees perform better, not simply become another tool that makes their work more complicated.

7. What Skills Will Employees Need in an AI-Powered Business?

Employees will need to continuously improve their skills.

One of the most important skills I would teach employees is how to research and learn independently.

AI technology is changing rapidly. A training program delivered today may become outdated as new tools, capabilities, and workflows emerge.

It is impossible to train employees on every AI tool they might encounter.

Instead, businesses should help employees develop the ability to:

• Research new AI technologies.
• Understand how AI tools work.
• Identify appropriate use cases.
• Ask better questions.
• Evaluate AI-generated outputs.
• Understand the limitations of AI.
• Apply AI to their existing responsibilities.

For example, an accountant who previously spent two days preparing a report could investigate how AI-assisted tools might reduce the time required.

The employee does not necessarily need to become a software developer.

They need to understand the problem, research available solutions, and learn how to use AI effectively.

Business owners need AI literacy too

I believe business owners should adopt a hybrid approach.

They should understand AI sufficiently to identify opportunities and solve smaller problems independently.

However, for complex implementations, they should work with technology consultants and software developers who understand AI, system integration, security, and business processes.

This combination reduces unnecessary dependence on external vendors while ensuring that complex systems are implemented properly.

8. Can One Person Run a Business That Previously Required a Team of 20?

I believe we are already moving in that direction.

AI tools are increasingly capable of assisting with functions that traditionally required different employees.

A business owner can use AI to assist with:

• Marketing.
• Sales research.
• Customer support.
• Accounting workflows.
• Software development.
• Business reporting.
• Administrative tasks.

A single person can coordinate multiple AI tools and agents, each supporting a different business function.

This creates an opportunity for smaller businesses to operate with fewer resources than would previously have been required.

However, there is an important distinction between automating tasks and running an entire business.

Businesses still require people to make strategic decisions, manage relationships, ensure quality, handle exceptions, and take responsibility.

AI can reduce the amount of human effort required, but it does not eliminate the need for effective business leadership.

The opportunity is to build leaner organizations that use AI to multiply the capabilities of their employees.

9. How Can Businesses Trust AI Agents With Critical Workflows?

Businesses should not give AI agents complete control over critical operations from the beginning.

I believe AI agents should be treated similarly to new employees.

When a business hires an intern or a new employee, it does not immediately assign them a critical task and walk away.

The employee is trained, their work is reviewed, feedback is provided, and their performance is evaluated over time.

The same approach should apply to AI agents.

A human-in-the-loop approach

Consider a business that wants to use AI to send marketing emails to 1,000 customers.

Instead of allowing the AI to send all the emails immediately, the workflow could operate as follows:

1. AI generates the email content.
2. AI prepares a small sample batch.
3. The samples are sent to a human reviewer through a tool such as WhatsApp or Telegram.
4. The reviewer checks the content, accuracy, and tone.
5. The reviewer approves or requests changes.
6. Once approved, the system sends the emails to the wider audience.

This approach allows businesses to benefit from AI’s speed while maintaining human control.

The same principle can apply to financial transactions, customer communications, HR decisions, and other sensitive processes.

Who is responsible when AI makes a mistake?

The business cannot simply blame the AI.

Businesses need to establish clear accountability for AI-assisted operations.

A human-in-the-loop workflow helps ensure that critical actions are reviewed and approved by someone with the appropriate authority.

As businesses deploy multiple AI agents, they will also need people responsible for supervising those agents, reviewing their performance, and managing risks.

This could create opportunities for roles such as AI operations managers.

The role would involve overseeing AI workflows, monitoring performance, ensuring compliance, and coordinating the interaction between employees and AI systems.

AI automation should not mean removing accountability from the business.

10. How Can Kenyan Businesses Adopt AI Without Expensive Digital Transformation Projects?

Kenyan businesses have an opportunity to adopt AI without necessarily following every traditional stage of digital transformation.

Consider the introduction of mobile phones in Kenya.

During the landline era, many households and businesses did not first establish a landline connection before adopting mobile phones.

Some moved directly from having no telephone access to using mobile phones.

This is an example of technological leapfrogging.

I believe businesses can apply a similar approach to AI.

A business that currently relies on paper records, Excel spreadsheets, and WhatsApp may not necessarily need to implement an expensive enterprise resource planning system before exploring AI.

It could potentially move directly toward a customized digital solution.

Three possible approaches

Approach 1: AI-assisted custom software

A business can work with a developer to create an application that addresses its most pressing operational problems.

AI-assisted development can potentially reduce the cost and time involved in building certain applications.

Approach 2: Open-source ERP systems

Businesses can explore open-source ERP platforms and work with technical consultants to customize them.

This may provide access to integrated business functions without the licensing costs associated with some proprietary systems.

However, implementation, hosting, customization, maintenance, and support still require investment.

Approach 3: AI-powered workflows

Businesses can automate specific processes using AI tools without replacing their entire technology infrastructure.

For example, they could automate customer inquiries, document processing, reporting, or marketing workflows.

The most appropriate approach depends on the business’s size, budget, technical requirements, and operational complexity.

The key lesson is that businesses should not assume they need an expensive technology project to begin adopting AI.

11. How Will AI Change the Software Development Industry?

AI is changing how software is created.

With AI-assisted development, developers can use natural language descriptions, AI coding tools, and automated testing to accelerate parts of the software development process.

This may reduce the time and cost required to develop certain applications.

As a result, traditional software development pricing models may face pressure.

However, I believe the greatest change will be in what software developers sell.

Historically, many developers have focused on delivering software applications.

In an AI-powered environment, developers will increasingly need to focus on delivering business solutions.

From software development to business outcomes

Consider payroll software.

A business does not necessarily want a payroll application for its own sake.

It wants employees to be paid accurately and on time, with the appropriate records and processes in place.

The business outcome is payroll management, not the software itself.

This distinction matters.

As software development becomes more accessible, businesses will increasingly expect technology providers to solve operational problems rather than simply deliver applications.

Developers will need to understand:

• Customer needs.
• Business processes.
• Operational challenges.
• System integration.
• AI implementation.
• Security.
• Distribution.
• Business outcomes.

I believe the strongest software companies will combine technical expertise with a deep understanding of their customers.

What becomes the competitive advantage?

If competitors can build similar software features using AI, the technology itself may become less of a differentiator.

The stronger competitive advantages will increasingly come from:

1. Understanding customer needs.
2. Delivering measurable business outcomes.
3. Building strong customer relationships.
4. Reaching customers through effective distribution.
5. Designing solutions around real operational problems.

The future of software development is not simply about writing code faster.

It is about solving business problems better.

12. What Is the Biggest Mistake Businesses Should Avoid When Adopting AI?

The biggest mistake is attempting to automate everything at once.

AI adoption should be a gradual process.

Businesses should identify their most pressing problems, implement a solution, evaluate its performance, and then move to the next opportunity.

For example, a business might begin by automating basic customer support inquiries.

After confirming that the system works reliably, it could explore automating reporting or marketing workflows.

This approach provides several benefits:

• Reduces implementation risk.
• Allows employees to learn gradually.
• Helps business owners understand AI capabilities.
• Makes it easier to measure results.
• Avoids unnecessary technology spending.
• Builds confidence in AI adoption.

Businesses should also avoid adopting AI simply because a tool is popular.

Every implementation should have a clear business objective.

The question should not be, “Which AI tool should we purchase?”

It should be, “What problem are we trying to solve, and is AI the best way to solve it?”

Questions fréquemment posées

Can AI automate an entire business?

AI can automate or assist with many business functions, including customer support, marketing, administration, reporting, and software development. However, businesses still need human leadership, judgment, accountability, and oversight. The realistic objective is to automate as much routine work as is appropriate while retaining human control over critical decisions.

Is custom AI software cheaper than SaaS subscriptions?

It depends on the business requirements and total cost of ownership. Custom software may reduce recurring licensing costs and provide better workflow integration, but it introduces development, hosting, maintenance, security, and support expenses. Businesses should compare the long-term costs of both approaches before making a decision.

Should small businesses implement AI before ERP systems?

Not necessarily. Some businesses may benefit from AI-powered workflows or customized applications before implementing a full ERP. Others may find that an open-source or commercial ERP is the most practical foundation. The right decision depends on the business’s operational needs and complexity.

Will AI replace software developers?

AI is likely to automate parts of software development and change how developers work. Developers who understand business problems, architecture, integration, security, and customer outcomes will have opportunities to deliver more value beyond writing code.

How should a business start using AI?

Begin with an AI audit. Analyze existing workflows, identify repetitive tasks, evaluate the risks, and select one practical problem to solve. Test the solution, measure the results, and expand gradually.

How can businesses protect sensitive data when using AI?

Businesses should evaluate the data-handling practices of AI providers and avoid exposing sensitive information unnecessarily. Depending on their requirements, they can consider private AI environments, locally hosted models, or hybrid systems that keep sensitive data within their own infrastructure. Human oversight and appropriate security controls remain important.

Conclusion: The Question Is No Longer Whether AI Will Change Your Business

AI is creating a fundamental shift in how businesses operate.

It is changing how employees work, how software is developed, how businesses manage their costs, and how they deliver services to customers.

For Kenyan businesses, the opportunity is particularly interesting. AI-assisted development, open-source software, and automated workflows may make it possible to adopt modern business technology without necessarily undertaking expensive digital transformation projects.

However, adopting AI successfully requires more than purchasing tools.

Businesses need to understand their operational problems, redesign their workflows, train employees, protect sensitive data, and establish appropriate human oversight.

My recommendation to business owners is straightforward: Start with an AI audit.

Identify the most pressing operational problem in your business. Determine whether AI can help solve it. Implement the solution step by step, measure the results, and continue improving.

The businesses that succeed will not necessarily be those with the most AI tools.

They will be those that understand their customers, redesign their operations intelligently, and use AI to deliver better outcomes.

The AI revolution is already underway. The question is whether your business is ready to adapt.