AI Data Search

Exploring large datasets no longer requires coding expertise. With the AI Data Search, both technical and non-technical users can now search and interact with complex data using plain language. There’s no need to write code or understand query syntax. Whether you’re an analyst, a product manager, or a sales rep, you can easily explore data, identify insights, and accelerate decisions.

All you need to do is enter a query in everyday language, and the Data Search interprets your intent, runs the appropriate search, and delivers results within seconds.

Here’s what sets it apart:

• No setup required. You don’t need to configure anything. Start chatting and get answers right away.

• Transparent logic. The search shows how it generated the results, so you can validate the output.

• Guided exploration. It offers suggestions to refine your queries or explore related paths when your request is too broad or returns limited data.

Write an effective prompt

Everything starts with your prompt. Follow these tips to help Data Search return the most relevant results:

• Be specific. Include details about the type of data you need (employees, companies, jobs), your target group, geographic scope, and phrase your prompt using terms that resemble real data fields and values whenever possible.

• Use precise criteria instead of broad adjectives. For example, rather than “biggest” or “top,” specify company size, revenue range, exact job titles, or required years of experience.

• Answer and clarify messages directly. When responding, be specific and concise – clear answers help Data Search narrow down the right data faster.

• Refine as you go. Review the results returned by Data Search, provide feedback, and adjust your prompt to improve accuracy.

• Leverage conversation context. Data Search reads all context in the conversation, so if results aren’t what you expect, ask it to explain its reasoning and review the clarifying messages it returns.

• Start fresh if needed. If you’re unhappy with the results or encounter an error, open a new message and try again with a revised prompt or choose from our suggestions.

Remove uncertainty

If your query is too vague or complex, the search will ask clarifying questions. For example:

“Do you mean companies by funding amount or by employee count?”

It also maintains conversational context so you can narrow or expand your search step-by-step:

“Now filter for companies in New York.”

This ensures you’re never stuck – even with partial or evolving requests.

Prompt examples

Here are several suitable prompts for you to find the targeted data:

• Find companies based in France with at least 50 active employees;

• Show all companies in LA that are working with Oracle;

• Find data engineers in the United States with at least 4 years of experience;

• List active product manager job postings in London.

Give feedback

Your feedback helps us improve Data Search. Here's how to share it:

1. Rate the response At the bottom of each Data Search response (below query), you'll find thumbs up and thumbs down icons.

   1. Click if you found Data Search returned relevant results to you. A feedback window will appear, where you can add additional feedback before submitting.

   2. Click if you found Data Search response unhelpful, inaccurate, or otherwise problematic. To provide more details, you can add them in the text field (optional).

2. Submit the feedback Click Submit to send your feedback. The feedback you submit includes your most recent search query and its search results.