Files
route-commerce/src/app/api/ai/stop-blast-advisor/route.ts
T

88 lines
3.5 KiB
TypeScript

import { NextRequest, NextResponse } from "next/server";
import { getAdminUser } from "@/lib/admin-permissions";
import { getAIClient } from "@/actions/integrations/ai-providers";
export async function POST(req: NextRequest) {
const adminUser = await getAdminUser();
if (!adminUser) {
return NextResponse.json({ error: "Unauthorized" }, { status: 401 });
}
if (!adminUser.can_manage_messages) {
return NextResponse.json({ error: "Forbidden" }, { status: 403 });
}
try {
const { brandId, stopId, stopName, stopDate, city, state, recentOrders, customerCount } = await req.json();
if (!brandId || !stopName) {
return NextResponse.json({ error: "Missing required fields" }, { status: 400 });
}
// Brand scoping
const effectiveBrandId = adminUser.brand_id ?? brandId ?? null;
if (!effectiveBrandId) {
return NextResponse.json({ error: "Brand ID required" }, { status: 400 });
}
if (adminUser.role === "brand_admin" && adminUser.brand_id !== brandId) {
return NextResponse.json({ error: "Not authorized for this brand" }, { status: 403 });
}
const aiResult = await getAIClient(effectiveBrandId);
if (!aiResult.client || !("client" in aiResult)) {
return NextResponse.json({ error: "error" in aiResult ? (aiResult as { error?: string }).error ?? "AI not configured" : "AI not configured" }, { status: 503 });
}
const SYSTEM = `You are a marketing strategist for a B2B produce wholesale platform.
Given a stop's order history and customer data, recommend the optimal stop blast message.
Return JSON with exact shape:
{
"timingRecommendation": "e.g., 'Send 2 days before the stop, between 9-11am for best open rates'",
"subjectLine": "email subject line",
"bodyPreview": "2-3 sentence email body preview",
"audienceSize": "estimated number of recipients",
"audienceRecommendation": "who should receive this (e.g., 'all pickup customers for this stop' or 'customers who ordered from this stop in last 90 days')",
"contentAngles": [
{
"angle": "e.g., 'Freshness angle'",
"reasoning": "why this angle works for this stop"
},
{
"angle": "e.g., 'Urgency/discount angle'",
"reasoning": "why this angle works"
}
],
"warnings": ["potential issue 1"]
}`;
const userMessage = `Brand: ${effectiveBrandId}
Stop: ${stopName}
Date: ${stopDate ?? "unknown"}
City: ${city ?? ""}, ${state ?? ""}
Recent Orders: ${JSON.stringify(recentOrders ?? [])}
Customer Count: ${customerCount ?? "unknown"}
Analyze this stop's context and order history to recommend the optimal stop blast message.
Return JSON with timingRecommendation, subjectLine, bodyPreview, audienceSize, audienceRecommendation, contentAngles array, and warnings.`;
// eslint-disable-next-line @typescript-eslint/no-explicit-any
const client = aiResult.client as { chat: { completions: { create: (opts: unknown) => Promise<{ choices: Array<{ message: { content: string } }> }> } } };
const response = await client.chat.completions.create({
model: aiResult.model,
messages: [
{ role: "system", content: SYSTEM },
{ role: "user", content: userMessage },
],
response_format: { type: "json_object" },
temperature: 0.4,
});
const content = response.choices[0]?.message?.content;
if (!content) {
return NextResponse.json({ error: "No response from AI" }, { status: 500 });
}
return NextResponse.json(JSON.parse(content));
} catch (err) {
return NextResponse.json({ error: "AI analysis failed" }, { status: 500 });
}
}