Beyond the Death of SaaS: The Real Possibility of SIer Extinction

After the “Death of SaaS,” Is the “Death of SIer” Next?

Anthropic has announced “Claude for Small Business,” an AI-powered business support service for small and medium-sized enterprises. The service features 15 different AI agents that can take over tasks such as customer support, data entry, and report generation.

It would be premature to dismiss this news as just another new AI service. Rather, this marks the beginning of a major tectonic shift that could fundamentally shake up the industry structure.

A recent article on Business + IT describes it as “After the ‘Death of SaaS,’ Is the ‘Death of SIer’ Next?” I strongly agree with this assessment. The traditional model where SaaS (Software as a Service) supported business efficiency is about to be dramatically transformed by the rise of AI agents.

And what lies ahead is the potential extinction of SIers (System Integrators). This is because AI agents have the potential to render the very business model of SIers—providing software implementation support—obsolete.

Business Transformation Brought by 15 AI Agents

The 15 AI agents offered by Claude for Small Business are specialized for specific tasks. For example, they include agents that automatically respond to customer inquiries, automate invoice processing, and generate data analysis reports.

These are not just simple chatbots. Each agent autonomously executes tasks within its granted permissions. It accesses necessary databases, outputs in the appropriate format, and asks for human confirmation when needed.

In my own company, I use three AIs—Claude, ChatGPT, and Grok—in combination, managing 32 agents. Specifically, I cover 29 business areas, including automated social media posting, WordPress article generation, video pipeline automation, contract review, and FX trading.

From this experience, I can say that AI agents are not “tools” but “team members.” While traditional SaaS provided “functions,” AI agents deliver “results.”

The Death of SaaS: Why AI Agents Are Replacing It

The essence of the SaaS business is providing specific business functions via the cloud. For example, Salesforce handles customer management, HubSpot handles marketing automation, and Slack handles internal communication.

However, AI agents can replace these across the board. This is because AI agents understand “what needs to be done” and can make their own decisions and act accordingly.

Let’s think about this concretely. Traditionally, a sales team would implement the following SaaS tools:

• Email distribution tool (e.g., Mailchimp)
• CRM (e.g., Salesforce)
• Scheduling tool (e.g., Calendly)
• Proposal creation tool (e.g., Canva)
• Contract management tool (e.g., DocuSign)

Each had to be contracted separately, integrated, and employees had to learn how to use them. Implementation costs ranged from a few hundred to several thousand dollars per month, and the setup period took weeks to months.

In contrast, an AI agent functions as a “comprehensive support from lead approach to contract signing.” It retrieves necessary information from databases, sends emails at optimal times, adjusts schedules, and automatically generates proposals.

In other words, the need to combine multiple SaaS tools disappears. This is why it’s called the “death of SaaS.”

The Death of SIer: Changing the Norms of System Development

The SIer business model involves hearing a company’s requirements, then designing, developing, and implementing a system. Large-scale projects can have budgets of hundreds of thousands of dollars and take months to years to build.

However, AI agents fundamentally change this process. This is because AI agents do not require “requirements definition” or “system design.”

For example, suppose a company wants to “automate responses to customer inquiries.” Traditionally, they would need to commission an SIer to build an inquiry management system or implement and customize a SaaS solution.

But with a service like Claude for Small Business, you can set up an AI agent for handling inquiries in minutes. All you need are the instructions (prompts) to give the agent and the database it will reference.

Let me share a personal experience. I used AI for legal negotiations in Malaysia. By feeding all emails, regulations, and laws into the AI for analysis and restructuring, I achieved faster, cheaper, and more accurate negotiations than hiring a lawyer.

This is a perfect example of building a “system” without an SIer. The AI played the role of “understanding requirements,” “designing,” and “implementing.”

Three Actions Business Leaders Should Take Now

To avoid being left behind by this change, business leaders should take the following three actions immediately.

1. Redesign Business Processes Assuming AI Agent Replacement

First, classify your company’s tasks into “tasks that can be delegated to AI agents” and “tasks that only humans can do.” Actively consider replacing delegable tasks with AI agents.

Specifically, it’s effective to start with tasks like these:

• Routine inquiry responses
• Data entry and aggregation
• Report generation
• Schedule coordination
• Initial contract review

In my own company, by delegating these tasks to AI agents, we generate value equivalent to approximately $50,000 per year at a cost of about $140 per month. That’s an ROI of roughly 2,989%.

2. Audit SaaS Contracts and Plan Migration to AI Agents

List the SaaS tools you currently subscribe to and evaluate whether their functions can be replaced by AI agents. In particular, prioritize reviewing the following types of SaaS:

• SaaS that offers only a single function
• SaaS with low employee usage frequency
• SaaS that is cumbersome to integrate with other tools
• SaaS with high monthly subscription fees

You don’t have to migrate all at once. I recommend starting with one task, confirming the results, and gradually expanding the scope.

3. Break Free from SIer Dependence and Establish Criteria for In-House Development

Identify system development projects currently outsourced to SIers and assess whether they can be replaced by AI agents. In particular, projects meeting the following conditions are worth considering for in-house development:

• Relatively simple requirements
• Development involving database integration
• Small number of users (a few dozen)
• Short development period (a few weeks)

In my experience, using tools like Claude Code or ChatGPT, even people with no programming experience can build simple systems. For example, an internal inquiry management system or inventory management system can be built in a few days with AI assistance.

Cost and Barriers to Implementation

While the detailed pricing for Claude for Small Business hasn’t been announced yet, it’s expected to be in a similar price range as existing plans like Claude Pro ($20/month) or Claude Team ($25/user/month).

If it’s around $50 per month, that would give you unlimited access to 15 AI agents for roughly $50/month. This could be cheaper than the cost of a single typical SaaS tool.

The barrier to implementation is also very low. Unlike traditional system implementation, there’s no need to create requirements definition documents or have endless meetings with vendors. You just tell the AI agent “what you want it to do,” and you can start using it immediately.

However, there are some points to note. To delegate tasks to an AI agent, you need the following preparations:

• Prepare the database the agent will reference
• Set permissions for the agent (which data it can access)
• Assign a human to check the agent’s output

These preparations are significantly lighter than traditional system implementation. Half a day should be sufficient.

Conclusion: How AI Agents Are Changing Business Norms

Anthropic’s Claude for Small Business is not just a new product launch. It’s a signal that the IT strategies of companies that have relied on SaaS and SIers are about to fundamentally change.

As a business leader, you should focus on these three points:

1. AI agents are “team members,” not “tools”
2. The time has come to break free from dependence on SaaS and SIers
3. Implementation costs are extremely low, and ROI is very high

Whether you view this change as “something for the future” or “an opportunity to act now” will significantly impact your company’s competitiveness in the years to come.

I have experienced the benefits of using AI agents in my own company. I encourage you to start by introducing an AI agent for just one task in your company. The impact will likely be greater than you imagine.