algo-dsl

Embed Recommendation Algorithms Language

Embed Studio DSL search, features, scoring, and ranking.

Install

npm install algo-dsl

Setup

import { StudioConfig, StudioV1 } from 'algo-dsl';

const config = new StudioConfig({ apiKey: 'YOUR_API_KEY' });
const mbd = new StudioV1({ config });

Usage

Set the target user for personalization

const polymarketWallet = '0x123...';
mbd.forUser("polymarket-wallets", polymarketWallet);

Generate candidates

Search your mbd indices using inclide/exclude filters and boosting options.

const candidates = await mbd.search()
  .index("polymarket-items")
  .includeVectors(true)
  .include()
  .numeric("volume_1wk", ">=", 10000)
  .exclude()
  .term("closed", true)
  .term("price_under05_or_over95", true)
  .boost()
  .groupBoost("polymarket-wallets", "ai_labels_med", polymarketWallet, "label", 1, 5, 10)
  .groupBoost("polymarket-wallets", "tags", polymarketWallet, "tag", 1, 5, 10)
  .execute();

Add candidates to current context

Attach the search results to the SDK so later steps can use them.

mbd.addCandidates(candidates);

Enrich your data

Fetch features (signals, metadata) for each candidate.

const features = await mbd.features("v1")
  .execute();
mbd.addFeatures(features);

Run predictive and reranking AI models

Score candidates with ML models for relevance or reranking.

const scores = await mbd.scoring()
  .model("/scoring/ranking_model/polymarket-rerank-v1")
  .execute();
mbd.addScores(scores, "rerank_polymkt1");

Combine all the data into final recommendations

Merge signals and produce the final ranked list with diversity and limits.

const ranking = await mbd.ranking()
  .sortingMethod('mix')
  .mix("topic_score", 'desc', 40)
  .mix("user_affinity_score", 'desc', 40)
  .mix("rerank_polymkt1", 'desc', 20)
  .diversity('semantic')
  .lambda(0.5)
  .horizon(20)
  .limitByField()
  .every(10)
  .limit("cluster_1", 1)
  .execute();
mbd.addRanking(ranking);

API

Method	Description
forUser(index, userId)	Set user context for personalization
search()	Build and run a search query
frequentValues(index, field, size?)	Fetch frequent values in an index/field (default size: 25)
addCandidates(array)	Add search hits to the current context
features(version)	Fetch features for candidates
addFeatures(result)	Attach features to candidates
scoring()	Run scoring/reranking models
addScores(result, key)	Attach model scores to candidates
ranking()	Produce final ranked recommendations
addRanking(result)	Attach ranking scores to candidates

Useful links

• Examples: https://github.com/ZKAI-Network/mbd_studio_demo
• Embed Studio webapp: https://api.mbd.xyz/v3/studio/frontend/