feat: migrate AI to Vercel AI SDK, add OpenAI provider, fix worklist

- Replace raw @anthropic-ai/sdk with Vercel AI SDK (generateText, tool, generateObject)
- Add provider abstraction (ai-provider.ts) — swap OpenAI/Anthropic via env var
- AI chat controller: dynamic KB from platform (clinics, packages, insurance), zero hardcoding
- AI enrichment service: use generateObject with Zod schema instead of manual JSON parsing
- Worklist: resolve agent name from platform currentUser API instead of JWT decode
- Worklist: fix GraphQL field names to match platform remapping (source, status, direction, etc.)
- Config: add AI_PROVIDER, AI_MODEL, OPENAI_API_KEY env vars

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

src/ai/ai-enrichment.service.ts

@@ -1,6 +1,9 @@
 import { Injectable, Logger } from '@nestjs/common';
 import { ConfigService } from '@nestjs/config';
-import Anthropic from '@anthropic-ai/sdk';
+import { generateObject } from 'ai';
+import type { LanguageModel } from 'ai';
+import { z } from 'zod';
+import { createAiModel } from './ai-provider';
 
 type LeadContext = {
     firstName?: string;
@@ -19,23 +22,25 @@ type EnrichmentResult = {
     aiSuggestedAction: string;
 };
 
+const enrichmentSchema = z.object({
+    aiSummary: z.string().describe('1-2 sentence summary of who this lead is and their history'),
+    aiSuggestedAction: z.string().describe('5-10 word suggested action for the agent'),
+});
+
 @Injectable()
 export class AiEnrichmentService {
     private readonly logger = new Logger(AiEnrichmentService.name);
-    private client: Anthropic | null = null;
+    private readonly aiModel: LanguageModel | null;
 
     constructor(private config: ConfigService) {
-        const apiKey = config.get<string>('ai.anthropicApiKey');
-        if (apiKey) {
-            this.client = new Anthropic({ apiKey });
-        } else {
-            this.logger.warn('ANTHROPIC_API_KEY not set — AI enrichment disabled, using fallback');
+        this.aiModel = createAiModel(config);
+        if (!this.aiModel) {
+            this.logger.warn('AI not configured — enrichment uses fallback');
         }
     }
 
     async enrichLead(lead: LeadContext): Promise<EnrichmentResult> {
-        // Fallback if no API key configured
-        if (!this.client) {
+        if (!this.aiModel) {
             return this.fallbackEnrichment(lead);
         }
 
@@ -48,7 +53,10 @@ export class AiEnrichmentService {
                 ? lead.activities.map(a => `- ${a.activityType}: ${a.summary}`).join('\n')
                 : 'No previous interactions';
 
-            const prompt = `You are an AI assistant for a hospital call center at Global Hospital.
+            const { object } = await generateObject({
+                model: this.aiModel!,
+                schema: enrichmentSchema,
+                prompt: `You are an AI assistant for a hospital call center.
 An inbound call is coming in from a lead. Summarize their history and suggest what the call center agent should do.
 
 Lead details:
@@ -60,25 +68,11 @@ Lead details:
 - Contact attempts: ${lead.contactAttempts ?? 0}
 
 Recent activity:
-${activitiesText}
-
-Respond ONLY with valid JSON (no markdown, no code blocks):
-{"aiSummary": "1-2 sentence summary of who this lead is and their history", "aiSuggestedAction": "5-10 word suggested action for the agent"}`;
-
-            const response = await this.client.messages.create({
-                model: 'claude-haiku-4-5-20251001',
-                max_tokens: 200,
-                messages: [{ role: 'user', content: prompt }],
+${activitiesText}`,
             });
 
-            const text = response.content[0].type === 'text' ? response.content[0].text : '';
-            const parsed = JSON.parse(text.trim());
-
             this.logger.log(`AI enrichment generated for lead ${lead.firstName} ${lead.lastName}`);
-            return {
-                aiSummary: parsed.aiSummary,
-                aiSuggestedAction: parsed.aiSuggestedAction,
-            };
+            return object;
         } catch (error) {
             this.logger.error(`AI enrichment failed: ${error}`);
             return this.fallbackEnrichment(lead);