Semantically Index your Data for Search, RAG & Entity Resolution

VectorLink is a semantic indexer, that enables you to use advanced AI techniques and OpenAI with your data and content.

Need help getting started? Talk to us.

Using OpenAI Vector Embeddings

VectorLink provides developers and non-technical users with semantic data and content management tools using vector embeddings.

It uses OpenAI to define the embeddings, and after an indexing request is made, your documents are sent to OpenAI which returns a list of float vectors in JSON.

To obtain good semantics, VectorLink creates high-quality text embeddings by defining a GraphQL query together with a Handlebars template rendering it as text so OpenAI can understand it.

After your data is indexed, you can ask the semantic index server questions about your data and the magic of the vector proximity produces semantic results.

				
					{
    "embedding": {
        "query": "query($id: ID){ People(id : $id) { birth_year, created, desc, edited, eye_color, gender, hair_colors, height, homeworld { label }, label, mass, skin_colors, species { label }, url } }",
        "template": "The person's name is {{label}}.{{#if desc}} They are described with the following synopsis: {{#each desc}} *{{this}} {{/each}}.{{/if}}{{#if gender}} Their gender is {{gender}}.{{/if}}{{#if hair_colors}} They have the following hair colours: {{hair_colors}}.{{/if}}{{#if mass}} They have a mass of {{mass}}.{{/if}}{{#if skin_colors}} Their skin colours are {{skin_colors}}.{{/if}}{{#if species}} Their species is {{species.label}}.{{/if}}{{#if homeworld}} Their homeworld is {{homeworld.label}}.{{/if}}"
    }
}

Using the People class in the Star Wars demo project, this example shows the GraphQL query and Handlebars template definition.

Consultancy Services

We provide consultancy services to help your development teams implement and manage VectorLink to leverage the full power of LLMs with your information. Please get in touch to discuss your needs. Our services include..

Semantic Search

Find answers to difficult questions. What did data look like a year ago, what products are impacted by a component shortage?

Entity Resolution

See what records refer to the same real-world thing and resolve these entities into a single record.

Retrieval-Augmented Generation (RAG)

RAG consultancy to incorporate your data with LLM parameter knowledge to build smarter AI tools.

Semantic AI Classification

Taxonomies, audiences, and categorisation with AI assistance to save time and resource.

VectorLink Semantic Uses

The ability to semantically index your data and leverage OpenAI provides the following features –

Semantic Search

Achieve precise and contextually relevant search results with vector embeddings.

Entity Resolution

Identify and eliminate duplicate content using intelligent entity resolution powered by vector embeddings

Similarity Search

Enhance content discovery and recommendation systems with advanced similarity search capabilities

Retrieval-Augmented Generation

Retrieval-augmented generation combines information retrieval of your business data with AI text generation.

Incremental Indexing

Keep up-to-date with real-time content changes through incremental indexing, ideal for CI workflows

VectorLink Sectors

Biotech & Pharmaceutical

Aid research through academic paper semantic search, clustering data points and test results, and finding similarities in datasets

Corporate Intelligence

Resolve and harmonize company or investor records

Manufacturing Supply Chain

Resolve data from many databases, cluster data to identify patterns, and apply semantic search for more accurate and context-aware search capabilities

E-commerce

Personalize product recommendations and improve search accuracy for online shoppers

Media and Publishing

Manage large volumes of articles, images, and videos efficiently, delivering targeted content

Knowledge Management

Organize and retrieve information from complex document repositories with ease

Digital Marketing

Optimize content discovery, recommendation engines, and targeted campaigns for effective customer engagement

Research and Development

Rapidly access relevant scientific articles and research papers to accelerate innovation

Financial Services

Fraud detection, customer identity matching, risk assessment, and due diligence are some ways financial services organizations can utilize VectorLink

4 Steps to Get Started

Add your OpenAI key to your team

Visit your profile page and copy and paste the key from OpenAI and turn indexing on.

Configure Vector Embeddings

Create the vector embeddings for the document classes you want to index.

Index Your Data

To begin indexing your data, create a change request and approve it. Indexing is versioned, so happens on a commit level.

Use VectorLink

Ask the semantic index server questions about your data and content to gain a deeper understanding of it.

Leverage AI's Potential

If you need to leverage the potential of your data and content using generative AI to solve difficult problems, our consultancy services can speed up the project. Enquire today to see about working with us.