Implement Integrated Inference

[austin-denoble]

Problem

The integrated inference API was part of the 2025-01 release, and the code has been generated for Go, but the interface has not been wired up for use in the client.

Solution

Implement Integrated Inference:

• Add CreateIndexForModel to Client struct.
• Add UpsertRecords and SearchRecords to IndexConnection struct.
• Add new types for working with integrated inference:
  • CreateIndexForModelRequest, CreateIndexForModelEmbed
  • IntegratedRecord
  • SearchRecordsRequest, SearchRecordsQuery, SearchRecordsRerank, SearchRecordsVector
  • Hit, SearchRecordsResponse, SearchUsage

Note: I've included some integration test refactoring in this PR. Was running into a lot of flakiness unrelated to this change.

Type of Change

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update
• Infrastructure change (CI configs, etc)
• Non-code change (docs, etc)
• None of the above: (explain here)

Test Plan

Following example demonstrates creating an integrated index for a specific model, upserting some records, and then searching against those records.

package main

import (
	"context"
	"fmt"
	"github.com/pinecone-io/go-pinecone/v3/pinecone"
	"log"
	"os"
)

func main() {
	ctx := context.Background()

	clientParams := pinecone.NewClientParams{
		ApiKey: os.Getenv("PINECONE_API_KEY"),
	}

	pc, err := pinecone.NewClient(clientParams)
	if err != nil {
		log.Fatalf("Failed to create Client: %v", err)
	} else {
		fmt.Println("Successfully created a new Client object!")
	}

	index, err := pc.CreateIndexForModel(ctx, &CreateIndexForModelRequest{
		Name:   "my-integrated-index",
		Cloud:  "aws",
		Region: "us-east-1",
		Embed: CreateIndexForModelEmbed{
			Model:    "multilingual-e5-large",
			FieldMap: map[string]interface{}{"text": "chunk_text"},
		},
	})

	if err != nil {
		log.Fatalf("Failed to create serverless integrated index: %v", err)
	} else {
		fmt.Printf("Successfully created serverless integrated index: %s", idx.Name)
	}

	idxConnection, err := pc.Index(pinecone.NewIndexConnParams{Host: index.Host, Namespace: "my-namespace"})

	records := []*IntegratedRecord{
			{
				"_id":        "rec1",
				"chunk_text": "Apple's first product, the Apple I, was released in 1976 and was hand-built by co-founder Steve Wozniak.",
				"category":   "product",
			},
			{
				"_id":        "rec2",
				"chunk_text": "Apples are a great source of dietary fiber, which supports digestion and helps maintain a healthy gut.",
				"category":   "nutrition",
			},
			{
				"_id":        "rec3",
				"chunk_text": "Apples originated in Central Asia and have been cultivated for thousands of years, with over 7,500 varieties available today.",
				"category":   "cultivation",
			},
			{
				"_id":        "rec4",
				"chunk_text": "In 2001, Apple released the iPod, which transformed the music industry by making portable music widely accessible.",
				"category":   "product",
			},
			{
				"_id":        "rec5",
				"chunk_text": "Apple went public in 1980, making history with one of the largest IPOs at that time.",
				"category":   "milestone",
			},
			{
				"_id":        "rec6",
				"chunk_text": "Rich in vitamin C and other antioxidants, apples contribute to immune health and may reduce the risk of chronic diseases.",
				"category":   "nutrition",
			},
			{
				"_id":        "rec7",
				"chunk_text": "Known for its design-forward products, Apple's branding and market strategy have greatly influenced the technology sector and popularized minimalist design worldwide.",
				"category":   "influence",
			},
			{
				"_id":        "rec8",
				"chunk_text": "The high fiber content in apples can also help regulate blood sugar levels, making them a favorable snack for people with diabetes.",
				"category":   "nutrition",
			},
		}

	err = idxConnection.UpsertRecords(ctx, records)
	if err != nil {
			log.Fatalf("Failed to upsert vectors. Error: %v", err)
	}

	res, err := idxConnection.SearchRecords(ctx, &SearchRecordsRequest{
			Query: SearchRecordsQuery{
				TopK: 5,
				Inputs: &map[string]interface{}{
					"text": "Disease prevention",
				},
			},
	})
	if err != nil {
			log.Fatalf("Failed to search records: %v", err)
	}
	fmt.Printf("Search results: %+v\n", res)
}

[austin-denoble]

nit: moved this block up so it lives with the other vector operations

[austin-denoble]

nit: realized all these helpers were being exported out of the package itself, they should be lowercase

[austin-denoble]

These helper functions are used in integration tests and are otherwise unrelated to the other changes in this PR. We were seeing a lot of test failures and I needed to add some additional logic to retry specific requests, and poll for freshness in a more targeted way.

[rohanshah18]

nit: Is this comment complete?

To create an index that accepts source text and converts it to vectors automatically using an embedding model hosted by Pinecone.

[austin-denoble]

Ah good catch, I think I lopped something off when editing.

[rohanshah18]

out of curiosity, are these changes manual or done as a part of the build process?

[austin-denoble]

I mean to add context for this in the PR, thanks for flagging.

This was manual, but it was due to PRs that have been put up by dependabot: #99

There was another one opened more recently to bump to 0.36.0, but that will also require bumping the base version of Go specified in the go.mod file, so I wanted to wait for a major release to do that.

[rohanshah18]

nit: not sure if you want to show the example on how to configure an index for a model

[austin-denoble]

Good point, I don't think we explicitly show that anywhere else. I can add an example.

[austin-denoble]

I actually missed implementing this for configure, thanks again. 👍

[rohanshah18]

Nice work, lgtm!!