Start Here: 3 Fast-Track Paths to Production Memory - Negentro Docs

Path 1: Store & Recall Memory in 2 Steps

Initialize your client with your live API key, store unstructured conversation facts, and perform multi-signal hybrid recall (dense vector + BM25 trigram fused with RRF_K=30).

TypeScript / Node.js — Minimal Companion Loop

import { PiyAPIClient } from '@piyapi/sdk';

const client = new PiyAPIClient({
  apiKey: process.env.PIYAPI_API_KEY || 'sk_live_prod_key_here',
});

// 1. Ingest user interaction fact
const memory = await client.memories.create({
  content: 'User prefers dark mode UI and works primarily with React and TypeScript.',
  tags: ['preferences', 'tech-stack'],
  metadata: { source: 'onboarding_chat', user_id: 'usr_9918' }
});

// 2. Perform hybrid search during subsequent chat turns
const relevant = await client.search.query({
  query: 'What frontend framework does the user prefer?',
  limit: 5,
  include_metadata: true
});

console.log('Recalled memory:', relevant.results[0].content);

Path 2: Strict Multi-Tenant Namespace Isolation

For SaaS applications with multi-tenant architectures, pass the X-Namespace-Prefix header on every request to guarantee zero scope leakage across client organizations.

cURL — Namespace Scoping

curl -X POST https://api.piyapi.cloud/api/v1/memories \
  -H "Authorization: Bearer sk_live_your_api_key_here" \
  -H "X-Namespace-Prefix: tenant_acme_corp" \
  -H "Content-Type: application/json" \
  -d '{
    "content": "Enterprise security policy mandates 90-day key rotations.",
    "tags": ["security", "policy"]
  }'

Path 3: Local Ollama & Air-Gapped Fallback

When running in privacy-sensitive or edge environments, use the BYOK routing engine or run local containers with pgvector and Ollama embeddings.