₦airaland Forum

[size=14]Salesforce in the AI Era: Should Businesses Prioritize AI, Integration, or Modernization?[/size]

Salesforce has become an important platform for businesses looking to improve customer relationships, streamline sales processes, and deliver better customer service. However, implementing a CRM is only one part of the journey. The real challenge often begins after implementation.

Disconnected enterprise applications, complex customizations, poor data visibility, low user adoption, and ongoing maintenance can prevent organizations from getting the full value of their Salesforce investment.

Now, with AI-powered capabilities such as Salesforce Agentforce gaining attention, businesses have another question to consider: should they introduce AI agents immediately, or first address the underlying issues in their CRM environment?

Here are five areas worth considering:

1. AI-Powered Agent Enablement

AI agents can support customer service workflows, help automate routine activities, and improve agent productivity. However, organizations need reliable data, appropriate access controls, and clear human escalation processes to make these capabilities effective.

2. Salesforce Implementation and Adoption

A successful implementation involves more than deploying Sales Cloud or Service Cloud. Businesses must align Salesforce with their processes, configure the right capabilities, support global rollouts, and make sure employees can use the platform effectively.

3. Integration with Enterprise Systems

Salesforce often needs to exchange information with ERP systems, third-party applications, and other cloud platforms. Well-designed integrations using APIs and suitable integration tools can reduce manual data handling and create a more consistent view of the customer.

4. Modernization and Product Engineering

As business requirements evolve, organizations may need to migrate from Salesforce Classic to Lightning, consolidate Salesforce environments, improve application performance, or develop custom products and applications. Assessment, migration planning, and automated testing can help manage these changes.

5. Managed Services and Continuous Improvement

Salesforce optimization does not end at go-live. Monitoring, technical support, issue resolution, and regular process improvements help organizations maintain performance as their requirements change.

For those exploring these areas, Datamatics has resources covering:

1. Salesforce Agentforce
2. Salesforce Implementation Services
3. Product Engineering Services
4. Modernization Services
5. Salesforce Managed Services
6. Salesforce Integration Services

The important thing is that not every organization needs the same approach. Some may benefit most from improving integrations, while others may need to modernize their existing environment before introducing AI-powered workflows.

What do you think?

If your organization uses Salesforce, what has been the biggest challenge: implementation, integration, user adoption, modernization, or ongoing maintenance? And would you prioritize AI-powered agents now, or first strengthen the existing CRM foundation?

Share your thoughts and experiences below.

Can Agentic AI really make underwriting decisions, or is it still just advanced automation?

I’ve been reading about how insurance companies are moving from traditional automation to something called Agentic AI, where AI agents can not only extract and analyze data but also coordinate multiple underwriting tasks and recommend decisions.

One example I came across was a Datamatics case study on AI-powered digital underwriting: https://www.datamatics.com/resources/case-studies/agentic-ai-powered-digital-underwriting

What caught my attention was the idea that the system could review documents, apply underwriting rules, assess risk signals, and support faster policy issuance rather than simply routing files between people.

My question to professionals in insurance, AI, or enterprise automation is:

At what point does a system become truly “agentic”? Is it when it can independently decide the next action, request missing information, and escalate exceptions on its own, or do you think these solutions are still heavily dependent on predefined workflows and human approval?

I’m also curious about the governance side. If an AI agent contributes to a risk assessment that later turns out to be wrong, should accountability remain entirely with the underwriter, or do insurers need a new model for AI-assisted decision ownership?

Would appreciate insights from anyone who has worked with underwriting automation, AI in insurance, or digital transformation projects.