Pinecone Client + Langchain OpenAIEmbeddings missing properties ts

[goku2130]

I have been trying to incorporate vector store containing OpenAIEmbeddings indexed on pinecone. I think this relates to the langchain ts interface. Is there any workaround?

Link to similar issue: https://community.pinecone.io/t/while-deploying-my-project-i-am-facing-issue-regrading-vectorstore-and-pineconeindex/3392

It would be great if someone could assist me.
Thanks

import { Redis } from '@upstash/redis';
import { OpenAIEmbeddings } from 'langchain/embeddings/openai';
import { PineconeClient } from '@pinecone-database/pinecone';
import { PineconeStore } from 'langchain/vectorstores/pinecone';

export class MemoryManager {
    private static instance: MemoryManager;
    private history: Redis;
    private vectorDBClient: PineconeClient;

    public constructor() {
        this.history = Redis.fromEnv();
        this.vectorDBClient = new PineconeClient();
    }

    public async init() {
        if (this.vectorDBClient instanceof PineconeClient) {
            await this.vectorDBClient.init({
                apiKey: process.env.PINECONE_API_KEY!,
                environment: process.env.PINECONE_ENVIRONMENT!
            });
        }
    }

    public async vectorSearch(
        recentChatHistory: string,
        companionFileName: string
    ) {
        const pineconeClient = <PineconeClient>this.vectorDBClient;

        const pineconeIndex = pineconeClient.Index(
            process.env.PINECONE_INDEX || ('' as string)
        );

        const vectorStore = await PineconeStore.fromExistingIndex(
            new OpenAIEmbeddings({ openAIApiKey: process.env.OPENAI_API_KEY }),
            { pineconeIndex }<<<<<< Error Type 'VectorOperationsApi' is missing the following properties from type >>>>>>>>>>>>>>>>>>>>>>'Index<RecordMetadata>': config, target, deleteAll, _deleteAll, and 10 more.ts(2740)
>>>>>>>>>>>>>>>>>>>>>>pinecone.d.ts(8, 5): 
        );

        const similarDocs = await vectorStore
            .similaritySearch(recentChatHistory, 3, { FileName })
            .catch((error) => {
                console.log('Failed to get vector seraech results: ', error);
            });

        return similarDocs;
    }
}

[dosubot[bot]]

🤖

Hello,

From the error message you provided, it seems like the PineconeStore.fromExistingIndex method is expecting an instance of Index<RecordMetadata> but it's receiving an instance of VectorOperationsApi.

The PineconeClient.Index method returns an instance of VectorOperationsApi which doesn't have the properties required by Index<RecordMetadata>. This is why you're seeing the error message.

To resolve this issue, you need to ensure that the pineconeIndex variable is an instance of Index<RecordMetadata>. Here's how you can do it:

const pineconeIndex = pineconeClient.Index(
    process.env.PINECONE_INDEX || ('' as string)
) as Index<RecordMetadata>;

This will tell TypeScript to treat the pineconeIndex variable as an instance of Index<RecordMetadata>, which should resolve the error.

Please note that this is a workaround and might not work if the actual Index<RecordMetadata> interface requires methods that are not available in VectorOperationsApi. In that case, you might need to extend the VectorOperationsApi class to include the missing methods.

I hope this helps! Let me know if you have any other questions.

Best,
Dosu

Sources

• examples/src/retrievers/pinecone_self_query.ts
• examples/src/indexes/vector_stores/pgvector_vectorstore/pgvector.ts

---

This response is meant to be useful and save you time. It is not meant to be a precise solution, but rather a starting point for your own research.

Help me be more useful! Please leave a 👍 if this is helpful and 👎 if it is irrelevant.

If you want to continue the conversation, start your reply with @dosu-bot.