With Kernel, Array deployed AI agents that work directly in Salesforce, with no need for additional tools or platforms.

First, Kernel's cleaning agent validates all accounts in Array's CRM. It fixes broken website links, removes duplicates, and maps each account to the correct LinkedIn profile.

Using this clean foundation, Kernel’s research agent maps accounts to Array's precise industry segments. Instead of broad categories like "Finance", accounts are classified into specific verticals such as "Personal Loans" or "Banks". Each classification comes with transparent reasoning and confidence levels.

Kernel’s prioritization agent then tiers accounts based on their potential fit. The model analyzes company growth metrics, funding history, business models, and team composition. Each tier assignment includes the key factors that influenced the decision.Most critically, Kernel’s estimation agent provides an estimate of the potential ARR for each account. The model combines customer counts, website traffic, app usage data, and industry benchmarks to estimate deal size potential specifically for Array. Like all Kernel data points, these predictions include detailed reasoning.

The entire system improves through rep feedback. When reps spot incorrect data, they submit quick notes in Salesforce. Kernel processes this feedback, updating the model and sharing insights with RevOps leadership during monthly reviews.