Getting Started

This guide will help you quickly get started with Orichain by demonstrating common use cases for each of its core modules.

---

Creating Embeddings with Orichain

You can generate embeddings from user queries using the EmbeddingModel class. Both synchronous and asynchronous variants are available.

import os
from dotenv import load_dotenv
from orichain.embeddings import EmbeddingModel  # Use AsyncEmbeddingModel for async usage

load_dotenv()

embedding_model = EmbeddingModel(
    model_name="text-embedding-ada-002",
    provider="OpenAI",
    api_key=os.getenv("OPENAI_KEY")
)

query_embedding = embedding_model(user_message="Hey there!")

---

Generating LLM Responses with Orichain

Orichain provides easy integration with large language models for text generation tasks. Here’s an example using the asynchronous AsyncLLM class.

from orichain.llm import AsyncLLM  # Use LLM for synchronous calls

llm = AsyncLLM(
    model_name="gpt-4.1-mini",
    provider="OpenAI",
    api_key=os.getenv("OPENAI_KEY")
)

response = await llm(
    user_message="Why is the sky blue?"
)

---

Retrieving Documents with Orichain Knowledge Base

The KnowledgeBase class allows you to retrieve relevant documents from a vector database like Pinecone.

from orichain.embeddings import EmbeddingModel
from orichain.knowledge_base import KnowledgeBase

embedding_model = EmbeddingModel(
    model_name="text-embedding-ada-002",
    provider="OpenAI",
    api_key=os.getenv("OPENAI_KEY")
)

knowledge_base = KnowledgeBase(
    vector_db_type="pinecone",
    api_key=os.getenv("PINECONE_KEY"),
    index_name="your index name",
    namespace="your namespace"
)

# Create query embeddings for retrieval
query_embedding = embedding_model(user_message="How to contact your customer support?")

# Retrieve relevant data chunks
retrieved_chunks = knowledge_base(
    user_message_vector=query_embedding,
    num_of_chunks=5,
)

# Alternatively, fetch documents by IDs
retrieved_chunks = knowledge_base.fetch(
    ids=["ID_1", "ID_2", "ID_3"]
)

---

Detecting Language with Orichain

You can detect the language of user queries using the LanguageDetection class.

from orichain.lang_detect import LanguageDetection

lang_detect = LanguageDetection(
    languages=["ENGLISH", "ARABIC"],
    min_words=2
)

user_language = lang_detect(user_message="هل يمكنك مساعدتي في استفساري؟")

---

Next Steps

Check out the Code Examples for more practical demonstrations on how to integrate Orichain into your applications.