Why should you care?

AgentNest product teams need a shared vocabulary before wiring OpenAI, Semantic Kernel, or RAG into CRM, ERP, and healthcare modules.

See it live — copy this example

Paste into an ASP.NET Core 8+ / AgentNest project, then run with dotnet run (set your API keys in user-secrets).

// AgentNest.Api/Models/AiInsight.cs
public sealed record AiInsight(string Topic, double Confidence, string Source);

public static class LeadScorer
{
    public static AiInsight Score(string companySize, string industry) =>
        new("lead-priority",
            companySize.Contains("Enterprise") ? 0.92 : 0.61,
            "rules+future-ml");
}

What happened?

• AiInsight is a tiny contract every AgentNest feature can return.
• LeadScorer shows deterministic scoring today — the same shape later holds model output from an injected IChatClient.

Practice next

1. Create AgentNest.Api with dotnet new webapi -n AgentNest.Api.
2. Add the AiInsight record and LeadScorer class under Models/.
3. Expose GET /api/leads/score?company=Enterprise via a minimal endpoint.
4. Add a third field RiskLevel to AiInsight and map it in LeadScorer.
5. Return a list of insights for multi-factor lead scoring.

Remember

AI in AgentNest means data-driven decisions exposed as typed API responses. Start with clear contracts like AiInsight before adding model calls. ASP.NET Core hosts AI logic as ordinary injectable services.

CRM lead triage

Sales ops asks AgentNest to flag high-value enterprise leads before reps open the record.

Outcome: A simple scorer ships in one sprint; the same DTO later swaps in GPT ranking.