code example
Build RAG Retrieval with altor-vec
What this pattern solves: Local chunk selection before prompt assembly for browser-side retrieval augmented generation.
RAG is only as good as the chunks it retrieves. When the corpus is small enough to ship, client-side retrieval can cut cost and latency by selecting context locally before any model call happens.
This is attractive for documentation copilots, product tours, and local-first assistants because the retrieval layer becomes instant and predictable. You only spend remote tokens on generation, not search.
Install
npm install altor-vecConcept explanation
In a rag retrieval workflow, users usually describe intent in their own words. That is why vector search works well here: each record is turned into an embedding, the embeddings are indexed once, and later queries retrieve the nearest semantic neighbors instead of relying only on exact tokens. In practice this means the interface can respond to paraphrases, shorthand, and partial descriptions far better than a literal-only search box.
The browser is often the right place to do this when the corpus is moderate in size and safe to ship. The instant benefit is lower latency. The architectural benefit is that you remove a whole search service from the request path. That matters for keystroke-heavy interactions, offline-capable apps, and product surfaces where search should feel like a UI primitive rather than a network round trip.
This page uses a deterministic embedding helper so the sample is runnable with only altor-vec installed. That keeps the example honest and easy to paste into a demo project. The same pattern works whether the final model runs remotely, in WebGPU, or in another on-device runtime. altor-vec simply provides the chunk shortlist.
Runnable JavaScript example
The following snippet indexes a small in-memory dataset, performs a semantic lookup for how do I rotate an API key, and prints the nearest matches. It uses the real altor-vec API, including init(), WasmSearchEngine.from_vectors(), and search().
import init, { WasmSearchEngine } from 'altor-vec';
const dims = 12;
const records = [
{
"title": "API key rotation",
"text": "Create a new key, update your environment, then revoke the old key after validation.",
"meta": "security"
},
{
"title": "Webhook retries",
"text": "Events retry with exponential backoff for up to twenty-four hours.",
"meta": "integrations"
},
{
"title": "Rate limit policy",
"text": "Per-token and per-project limits plus guidance for queueing requests.",
"meta": "platform"
},
{
"title": "Role management",
"text": "Owners, admins, and viewers have different workspace permissions.",
"meta": "auth"
},
{
"title": "Data export guide",
"text": "Export logs, traces, and usage reports from the admin console.",
"meta": "data"
},
{
"title": "Audit events",
"text": "Security-sensitive actions are recorded in the audit event stream.",
"meta": "security"
}
];
function embedText(text) {
const vector = new Float32Array(dims);
for (const token of text.toLowerCase().split(/[^a-z0-9]+/).filter(Boolean)) {
let hash = 2166136261;
for (const char of token) {
hash = Math.imul(hash ^ char.charCodeAt(0), 16777619);
}
const slot = Math.abs(hash) % dims;
vector[slot] += 1;
vector[(slot + token.length) % dims] += token.length / 10;
}
const magnitude = Math.hypot(...vector) || 1;
return Array.from(vector, (value) => value / magnitude);
}
async function main() {
await init();
const flat = new Float32Array(
records.flatMap((record) => embedText(`${record.title} ${record.text} ${record.meta}`))
);
const engine = WasmSearchEngine.from_vectors(flat, dims, 16, 200, 64);
const hits = JSON.parse(engine.search(new Float32Array(embedText('how do I rotate an API key')), 4));
const results = hits.map(([id, distance]) => ({
...records[id],
similarity: Number((1 - distance).toFixed(3)),
}));
console.table(results);
engine.free();
}
main();React component version
The React version keeps the same index build but wires it into component state so the UI can query on input changes. That is usually how teams introduce semantic retrieval into an existing product: initialize once, keep the engine in memory, and map nearest-neighbor hits back to the original records.
import { useEffect, useState } from 'react';
import init, { WasmSearchEngine } from 'altor-vec';
const dims = 12;
const records = [
{
"title": "API key rotation",
"text": "Create a new key, update your environment, then revoke the old key after validation.",
"meta": "security"
},
{
"title": "Webhook retries",
"text": "Events retry with exponential backoff for up to twenty-four hours.",
"meta": "integrations"
},
{
"title": "Rate limit policy",
"text": "Per-token and per-project limits plus guidance for queueing requests.",
"meta": "platform"
},
{
"title": "Role management",
"text": "Owners, admins, and viewers have different workspace permissions.",
"meta": "auth"
},
{
"title": "Data export guide",
"text": "Export logs, traces, and usage reports from the admin console.",
"meta": "data"
},
{
"title": "Audit events",
"text": "Security-sensitive actions are recorded in the audit event stream.",
"meta": "security"
}
];
function embedText(text) {
const vector = new Float32Array(dims);
for (const token of text.toLowerCase().split(/[^a-z0-9]+/).filter(Boolean)) {
let hash = 2166136261;
for (const char of token) {
hash = Math.imul(hash ^ char.charCodeAt(0), 16777619);
}
const slot = Math.abs(hash) % dims;
vector[slot] += 1;
vector[(slot + token.length) % dims] += token.length / 10;
}
const magnitude = Math.hypot(...vector) || 1;
return Array.from(vector, (value) => value / magnitude);
}
export function RagRetrievalExample() {
const [engine, setEngine] = useState(null);
const [query, setQuery] = useState('');
const [results, setResults] = useState([]);
useEffect(() => {
let cancelled = false;
let instance;
(async () => {
await init();
const flat = new Float32Array(
records.flatMap((record) => embedText(`${record.title} ${record.text} ${record.meta}`))
);
instance = WasmSearchEngine.from_vectors(flat, dims, 16, 200, 64);
if (!cancelled) setEngine(instance);
})();
return () => {
cancelled = true;
instance?.free();
};
}, []);
useEffect(() => {
if (!engine || query.trim().length < 2) {
setResults([]);
return;
}
const hits = JSON.parse(engine.search(new Float32Array(embedText(query)), 5));
setResults(
hits.map(([id, distance]) => ({
...records[id],
similarity: Number((1 - distance).toFixed(3)),
}))
);
}, [engine, query]);
return (
<section>
<input
value={query}
onChange={(event) => setQuery(event.target.value)}
placeholder="Ask a knowledge question"
/>
<ul>
{results.map((result) => (
<li key={result.title}>
<strong>{result.title}</strong> — {result.meta} (score {result.similarity})
</li>
))}
</ul>
</section>
);
}How this example works
The pattern has three moving parts. First, you choose what text represents each record: title, description, metadata, or a chunk of content. Second, you turn that text into vectors and flatten them into one Float32Array. Third, you build the HNSW graph and query it with a vector created from the user input. The library returns nearest-neighbor IDs and distances, and your app decides how to display or post-process them.
Because the retrieval step is approximate nearest-neighbor search, it stays fast even as the dataset grows beyond trivial linear scans. The most important quality lever is still the embedding model. Better vectors usually matter more than micro-optimizing ANN parameters, so teams should benchmark retrieval quality on real user phrasing before shipping the experience widely.
When to use this pattern
This is a practical fit when the search corpus is small to medium, shipped with the app, and searched frequently enough that backend latency would be noticeable. Common examples include docs portals, embedded support widgets, local-first assistants, and curated catalogs.
- Docs copilots
- Embedded support chat
- Product assistants
- On-device RAG prototypes
Limitations
RAG quality still depends on chunking, prompt assembly, and source freshness. For large private corpora or rapidly changing knowledge bases, server-side retrieval is often the safer choice.
Be especially careful about corpus size, update frequency, and data sensitivity. Browser vector search is excellent when those three constraints are favorable, but it is not the right answer when the dataset is huge, private, or changing constantly for every user.
FAQ
Can altor-vec be the retrieval layer for browser RAG?
Yes. That is a natural fit when the corpus is small enough to ship and you want fast local chunk recall.
Does this replace the language model?
No. It only retrieves relevant context. You still need a generation model or answer synthesizer downstream.
When should retrieval move back to the server?
When the corpus is large, access-controlled, frequently updated, or too sensitive to send to the browser.
Get started: npm install altor-vec · GitHub