LLM observability, agent memory, and product history — one event log.

What TES does

History, understanding, intelligence — in that order

Three layers built on one event spine. Each layer compounds the value of the one below it.

"Event sourcing provides a complete audit log of changes, which is crucial for understanding the behaviour of complex systems and for debugging."

— Martin Fowler, Event Sourcing, martinfowler.com. TES builds on this pattern with AI enrichment and vector search at every layer.

How it works

Emit, enrich, query — three steps for any data source

Whether it's an AI agent, a legacy ERP, or a warehouse scanner — the same three-step flow captures everything.

// From an AI agent
await tes.emit("agent_session.action_taken", {
  session_id: "ses_7f2a",
  action: "purchase",
  agent_id: "agent_007",
});

// Or a legacy system
await tes.emit("order.created", {
  order_id: "PO-2024-8891",
  source: "oracle_erp",
});

// After uploading an image, TES auto-enriches:
// → brand: "Levi's", condition: "good"
// → market_price: { low: 24, mid: 38, high: 55 }
// → embedding: [0.021, -0.043, ...] // 1024 dims
// → taxonomy: "Clothing > Denim > Jeans"

// Semantic search
const results = await tes.search({
  query: "vintage leather jacket",
  min_score: 0.7,
});

// Full audit trail
const history = await tes.project("thing", "thing_abc");
// → { stage: "listed", events: [...] }

Modules

Install intelligence onto your system

Each module extends what your system can understand and do — running on the same shared history. No duplication, no reprocessing.

Integration paths

Three ways to connect

Stream from a legacy database, connect an AI agent, or integrate with your existing stack — TES meets you where you are.

import { TESClient } from
  "@pentatonic/ai-agent-sdk";

const tes = new TESClient({
  apiKey: process.env.TES_API_KEY,
});

// Auto-track every LLM call
const ai = tes.wrap(new Anthropic());

const res = await ai.messages.create({
  model: "claude-sonnet-4-20250514",
  messages: [{ role: "user",
    content: "Find similar items" }],
});

{
  "mcpServers": {
    "tes": {
      "command": "npx",
      "args": [
        "-y",
        "@anthropic-ai/mcp-remote",
        "https://tes.pentatonic.com/sse"
      ]
    }
  }
}

query {
  searchThings(input: {
    query: "vintage leather jacket"
    min_score: 0.7
  }) {
    items {
      score
      thing { name vision { brand } }
    }
  }
}