A Fully Loaded SDK
For Next-Gen AI Search
Retrivora hides the complexity of setting up custom vectors, token indexes, embeddings, and chat pipelines under a single unified API.
Built for every layer of the stack
Swap providers without rewriting a single line of business logic.
Vector Store
Universal support for Pinecone, PGVector, MongoDB, Milvus, Qdrant, and more.
Embeddings
Seamlessly switch between OpenAI, Ollama, or custom embedding providers.
LLM Orchestration
Optimized inference across OpenAI, Anthropic, Gemini, and local LLMs.
Everything you need to build
From ingestion to inference — Retrivora handles the full RAG lifecycle.
Pluggable Vector Databases
Seamlessly switch between Pinecone, MongoDB, pgvector, Milvus, Qdrant, ChromaDB, and Weaviate, or implement custom database adapters (e.g. Astra DB, Elasticsearch) on the Enterprise plan.
LLM Provider Independence
Connect to OpenAI, Anthropic, Google Gemini, Ollama (Local), or any custom OpenAI-compatible API gateway instantly.
Client-Safe & Secure
Keep API keys and config protected on your server while exposing clean, secure endpoints for frontend widgets.
Pre-styled UI Components
Drop-in React components — ChatWidget, ChatWindow, DocumentUpload — with fluid animations and responsive mobile layouts.
Real-time Observability
Track latency breakdowns, token consumption, and active project requests with the built-in system telemetry stream.
Custom Ingestion Pipelines
Automatically chunk, parse, and upload PDF, TXT, or markdown files into your vector index using metadata tagging.
Ready to build your AI-powered app?
Install the SDK in 30 seconds. Your first RAG app in under 5 minutes.