Semantic Index Kit

Vector-based semantic search with hybrid retrieval.

0/6 documents indexed

Configuration

Semantic Indexing

arxiv_2301.07041.pdf47 chunks
1536d
anthropic_safety_blog.md23 chunks
1536d
nature_ml_survey.pdf89 chunks
1536d
techcrunch_ai_policy.html15 chunks
1536d
arxiv_2212.08073.pdf62 chunks
1536d
wikipedia_llm_evaluation34 chunks
1536d

Integration Code

import { createKnowledgeBase } from 'agent-tools-kit/knowledge'

const kb = createKnowledgeBase({ chunking: 'semantic', embedModel: 'text-embedding-3-small' })
await kb.ingest('./documents/')
const results = await kb.search('query text', { topK: 5 })