Latest developments

Hi guys, we're back again! This month, we have two highlights and some smaller improvements on the app functionality and enrichments.

Let’s talk about the details!

The Highlights of April

Huge Improvement: Custom Deduplication Rules With AI

Remember the AI Editing feature we dropped in January?

We just built the same functionality for the Deduplication Suite. It’s simple to use, yet powerful, and just like AI Editing, the only thing you need to do is to imagine what you want and tell our AI about it.

What It Does

AI Editing in the Deduplication Suite is basically the same thing as in the spreadsheet editor.

It allows you to edit, create, and transform your deduplicated rows with a single prompt.

You could, for example:

• Label properties - Ex: create a “Duplicate” checkbox and mark it for all items in a group.
• Edit data without deleting records - Ex: Copy the most common value for /Property to every item in the group. Don't delete items.
• Skip Groups - Skip groups with more than 2 duplicates. Then...
• Complex master item rule - Select the item with the longest text in {Property} as the master item. Delete all other items.

Why It Matters

Before, you would have to choose between merge or delete, and had some options to choose a master item, but you were still limited to a few options. Now you’re not limited anymore, just tell our AI Assistant what you want to do and it will do it.

After all, it’s you who knows best what to do with your data.

How to Use It

1. Go to Clean ⇒ Duplicates Finder
2. Run a duplicates check
3. Click on AI Editing
4. Imagine anything you want to do with your data
5. Write a prompt explaining it to the AI
6. Wait 10-20 seconds while the AI is writing the script
7. Look at the preview, confirm, or write a follow-up prompt to improve the outcome

Related Guides

• How to Unify Field Values in Duplicates (Without Merging)
• How to Sum Numeric Values When Merging Duplicates

Okay, now to highlight number two!

New Source: AI Agent - Site Scraper

This scraper is the first no-headache option to scrape data from any website; in fact, it makes scraping so easy that it becomes almost too easy. The only thing you do is: Prompting

What It Does

It visits websites, paginates them, understands the content, and extracts the data you need.

Here are some examples of what our AI Scraping Agent can do:

• Visit directory X and extract all listings from it
• Extract products of an Amazon search
• Scrape Shopify Stores
• … and much more

Why It Matters

If scraping isn’t your profession or hobby, then you shouldn’t be required to waste hours learning it. However, that’s what most of the scraping tools push you to do. This scraper is different.

With Datablist’s AI Scraping Agent, you can scrape websites without:

• Writing code
• Configuring unstable API’s
• Click-and-Point on website elements

Or losing your mind watching bad tutorials…..

The only thing you have to do: Tell an AI agent what to do (prompting)

How to Use It

• Create a collection
• Click See all sources, scroll down, select AI Agent - Site Scraper
• Paste a URL and write a prompt (or choose one of our templates)
• Configure output properties
• Set a limit
• That’s it.

Related Guides:

• How To Scrape a Directory
• How To Scrape a Shopify Store
• How To Scrape Zillow Properties
• How To Scrape AirBnB Properties
• How To Scrape Businesses From Yellow Pages
• How To Scrape Real Estate Agents From Zillow

Improvements

LinkedIn People Scraper

You can now choose to either 3, 5, or 10 past experiences when scraping a LinkedIn profile.

Why we did it: One of our customers was dealing with the accuracy of their lead lists.

The problem: LinkedIn and other B2B databases are not good at filtering, so sometimes when they were searching for CEO @ Company, e.g., Apple, they got results for people who are CEO at another company, and worked simultaneously at Apple.

The fix: By getting their complete employment history, they could run the Claude AI processor on their lead list and double-check if the prospect is really working at their target account or not.

Cleaning Preview

You can now see what changes you’re going to make when using the Bulk Edit or other cleaning features.

Why we did it: It makes it easier to understand what you're doing before committing to changes.

That’s it folks, until next time!

If you want us to build something for you, PITCH ME HERE 👈🏽

Hi everyone, March was huge, exciting, and funny! We shipped a lot of stuff.

Here’s a quick summary of what we did for you in March:

• Released a new out-of-the-box feature – Run Status column
• Integrated our first CRM - Datablist is now connected with PipeDrive
• Shipped a new data source - Instagram follower scraper
• Integrated 5 new LLMs - Claude, Perplexity, Gemini, DeepSeek, and Mistral
• Updated the LinkedIn scraper - Scraping and converting now: LinkedIn URLs and Sales Navigator IDs

The Highlights of March

Run Status Column – Get More Context

Do you remember a moment where having more context on what’s happening with your enrichments ever hurt you? Me neither.

How Was It Before

Before we implemented this, you had no chance to know what was happening behind the scenes. For example:

• The costs of each row for usage-based enrichments
• The types of errors
• What was causing the errors
• Where the data inconsistencies were

And as many of you were a bit frustrated since you were left in the dark about what's happening with your enrichment runs, we designed this feature to be insightful and easy to grasp.

Why This Matters

The Run Status Column gives you:

• Transparency and efficiency in your workflow.
• Instant visibility into what's happening with each enrichment

Which allows you to:

• Re-run enrichments with 3 mouse clicks
• Understand issues & errors faster
• Optimize your processes
• ….

By understanding where and why things might go wrong, you can make more data-driven decisions to improve your enrichment strategy.

These are the statuses you might see:

• Success & number of credits consumed
• Invalid data
• Missing data
• Rate limit notifications
• HTTP notifications

What You Can Do With It

• Reconfigure enrichments more easily
• See how many credits each row consumed
• Get more details on errors
• Fix prompting errors faster
• Re-run enrichments easier

First CRM Integration: PipeDrive CRM

If you use PipeDrive, you probably know how often duplicates can end up in your list, and you have to manually remove them after you try it with PipeDrive's native deduplication features, and it doesn’t work.

And it's not only PipeDrive; most CRMs have a big problem with data quality features. Yes, they all have a built-in duplicate finder, but this comes with huge limitations, such as:

• You can't search for duplicates across different Pipelines.
• You can only delete duplicates, not merge them.
• Duplicate searching is limited to some properties—it won't work with websites, email, or phone numbers.s
• The duplicate detection uses exact matching only, so it misses records with simple typos.
• There's no merge history or changelog to track changes.
• You have no control over which records to keep or delete, risking important data loss.
• There's no way to process duplicates in bulk.

Not trying to say that PipeDrive is bad. I am just saying that its deduplication features are limited because they have another focus, which is why we built this integration for you.

How It Works

• Import data from Pipedrive automatically.
• Clean the data inside of Datablist.
• Merge records in Pipedrive by ID (Merge one ID with another).
• Update and sync cleaned data with Pipedrive by copy-pasting your API key.

Related Resources

How to deduplicate and clean a PipeDrive CRM

New Features & Improvements

New Data Source: Instagram Follower Scraper

We had a user on a call asking for a way of scraping Instagram followers to monitor competitors and search for patterns across their followers, and since we want to help with the things that matter to our community, we were excited to build this feature quickly.

Want us to build something for you? Tell us about it here!

How It Works

• Give us the profiles
• Set a limit of followers you want to scrape

That’s it. Datablist does the rest

How To Use It

• Create a collection
• Click on “See all sources.”
• Select “Instagram follower scraper.”
• Paste a list of line-separated profiles.
• Set a limit on the number of followers to scrape
• Click on “Continue to outputs configuration.”
• Click on the ⊕ icons to add a new column for each output
• Click on “Import now.”

Related Guide: How To Scrape Instagram Followers

New Processors: Expanded LLM Choice

First days of March, sunny day in Nantes, Florian receives an email from a semi-upset customer saying: “I don’t want to be limited in my choice of the LLM, why don’t you integrate other models too?”

We heard that feedback and promptly expanded our AI capabilities by integrating five new powerful language models.

How Was It Before?

Our previous LLM options were limited to OpenAI models, which meant:

• Fewer choices for specific use cases
• Less flexibility in model selection
• Limited optimization options

Why This Matters

With Claude, Perplexity, Gemini, DeepSeek, and Mistral now available, you have more options to choose the right AI model for your specific needs, ensuring better results and more efficient processing.

What You Can Do With It

• Choose the best model for your specific use case
• Optimize for cost vs performance
• Access cutting-edge AI capabilities
• Compare results across different models

LinkedIn Profile Scraper - Normalize URLs & More Input Formats

Our LinkedIn Scraper now also allows Sales Navigator IDs as input and gives the normalized LinkedIn URL as output!

Some LinkedIn sequencer and message personalization tools don’t allow Sales Navigator URLs as input, which makes it impossible for some people to proceed with their workflows, so we made an update to our LinkedIn scraper

Why We Did It

Funny story, if we think about it for a moment: Someone who wasn’t even a customer needed to convert Sales Navigator IDs to LinkedIn Profile URLs, so we started looking for solutions.

2 Days later, we found the solution, implemented it, and reached out to that prospect, telling him, “Yo {{first_name}}, we found a solution and implemented it for you. Wanna try it out?

This was his answer: “Yo Habib, thanks so much, but I've found a solution.”

Do we regret it? No, quite the opposite, since other users also find their use in this improvement.

Related Guide: How To Convert LinkedIn Sales Navigator IDs to LinkedIn Profile URLs

That’s it, see y’all next time!

P.S. If you have any feature ideas you'd like us to build for you PITCH ME HERE ⟸ ⟸

Hi everyone! While this month's updates are modest, no big highlights, but they bring very important improvements.

Here's what we've been working on:

• Source configuration saving
• See prompt history in AI Editing

Though February's release may seem lighter on new features, sources, and enrichments compared to previous months, we have some exciting updates coming in March!

Let's look at these improvements in detail!

Improvements

Source Configuration: See and Re-Run From Saved Enrichments

From now on, you can access and re-run the data imports you previously ran through 4 simple mouse clicks, yes, just 4 clicks, no joke.

How It Was Before?

Being completely honest, it was not intuitive and a bit frustrating to re-run a source. Once you have already imported data into your collection. You'd have to:

1. Go to Imports
2. Choose the data source again
3. Reconfigure and run it.

Too many steps for a simple task. So we fixed it to make it easier for you

The Changes We Made

Every time you run a source now it gets saved in the Saved Enrichments

How To Use It

• Click on the Purple Blitz icon in your column bar
• Re-run the source

Simple, effective, fast.

How This Will Make Your Life Easier

• You can restart sources faster
• You can restart without needing to reconfigure the source from scratch
• You get more done: easier and faster.

AI Editing: See Prompt History of Run Scripts

The AI Editing feature is a powerful way of manipulating and editing your data at scale without needing to know how to write functions, since our AI will do it for you.

But there was a problem with it: no prompt would be saved.

As we introduced this feature, the feedback was nice, but after a few days, the first of you started telling us that you don’t want to always be rewriting prompts.

We heard it, we fixed it!

How It Was Before

• You configure a prompt
• You run the script that the AI wrote for you
• You do the same again and again, each time you get new data

This was annoying you, so we fixed it.

The Changes We Made

Each prompt you write gets saved now in a prompt history, allowing for easier re-runs

How This Will Make Your Life Easier

• You can re-run your workflow with 4 mouse clicks (yes, literally 4 clicks)
• No need to spend time writing a prompt again

Let me know what you think about the improvements and this "4 mouse clicks" thing (this wasn't planned, and my favorite number isn't 4 either; it's 1)

That’s it, folks!

Send us feedback ⇒ ⇒ Habib’s LinkedIn

Hi guys, we are starting this year with two highlights. Both are unmatched and will bring you a lot of value; this is a promise.

1. Job listings scraper - Scrapes job listings from 19 different boards simultaneously
2. AI Editing - Allows you to do complex data manipulation tasks, which otherwise would require code, with a simple prompt.

Now, to the details!

The Highlight of January

Job Postings Search – Scrape 19 Job Boards at Once

After having already released the Indeed jobs scraper and the LinkedIn jobs scraper, we saw how high the demand is for job market data, so we implemented this scraper, which delivers you data from the biggest job boards on the planet, including:

1. Indeed

2. LinkedIn

3. Glassdoor

4. Naukri.com

5. AngelList

6. InfoJobs

7. Tecnoempleo

8. Startup Jobs

9. SimplyHired

...and 10 others

Here’s what makes this scraper different from all the other job scrapers on the market:

• Scrapes fresh and up-to-date job postings
• Allows to search for keywords or phrases in job descriptions (awesome, isn’t it?)
• Global coverage of the labor market (195 Countries)
• Includes company information
• Includes hiring manager information

How It Works

You choose the starting point from the two following options:

• Scrape job listings of companies you've already added to your Datablist collection
• Start search from scratch using job titles, keywords, industries, funding stage, and 10 more filters

Datablist will then scrape the job listings that match your search and give you the results we find across those 19 job boards.

How To Use It

• Create a collection and select “Job Offers Search.”
• Map a collection to the search or start from scratch
• Use the 14 different filters to narrow down your search
• Click on “Continue to outputs configuration.”
• Click on the ⊕ icons to add a new column for each output
• Click on “Import now” to start scraping

New Feature: AI Editing

Florian is really excited about this one (and so am I)

Here’s why you should be excited too!

Datablist has a hidden strength, which is using JavaScript to manipulate data in any way you want it to be. Until now, this has been kept only for those who know how to write JavaScript scripts.

AI Editing brings the same power to non-technical folks using plain English instead of code.

How It Works

The functionality of this feature is pretty simple, as we want it to be.

Imagine having Claude, Gemini, or ChatGPT sitting in your spreadsheet, where you could just tell it what to do, because it’s exactly like this. Here’s how you could collaborate with our AI:

• You imagine anything you want to do with your data
• You write a prompt explaining it to the AI
• You wait 10-20 seconds while the AI is writing the script
• You look at the preview, confirm, or write a follow-up prompt to improve the outcome

Regardless of whether it's a certain structure, edit, system, or specific format you want to have, just tell your AI assistant about it, and it will do it for you.

What does that mean for you?

• You can use plain English to edit your data.
• You can build scoring systems with a single command.
• You can clean and format your data with simple prompts.
• … and so much more.

How To Use It

• Click on "Edit" in the Datablist top menu
• Select "AI Editing"
• Type your prompt
• Click on "Generate"
• Review the changes and click on "Run Script" to apply them, or click "Improve Prompt" to improve your prompt

Use Cases

• Format phone numbers
• Clean company names
• Build an account scoring system
• Capitalize words

Related Resources

Watch me build an AI scoring system with a simple prompt

That's it, folks, see y’all next time!

P.S. If you want us to build something for you, PITCH ME HERE 👈🏽

Hi folks, we are wrapping the year up with a huge UI update + two new sources.

Let’s get into it!

The Highlight of December

Preview Mode for Merging Duplicate Groups

This improvement in our preview mode is a big step towards making data handling and automation easy and accessible for everyone; it doesn't just add simplicity but makes it visual.

How It Was Before

Before we implemented this, the preview mode was not only harder to understand but also visually confusing when trying to understand how the data would be merged.

The preview mode wasn't intuitive and made it difficult for users to confidently make merging decisions. Additionally, there was no clear indication of which record would be the master record in the merging process.

The Changes We Made

• We added deeper and clearer descriptions and labels to every part
• We grouped the settings of the conflicting properties and separated them visually
• We separated the master item selection configuration visually
• We added labels to the merging preview of each duplicate group
• We added colors that highlight the master and secondary items
• We added an action to remove an item from the duplicate group

How This is Going To Make Your Life Easier

• Clear Visual Understanding: Thanks to our new color-coding and improved layout, you'll instantly see which records are being merged and how they fit together
• Reduced Error Risk: We've added better labels and grouping to make sure you don't accidentally merge the wrong records or pick the wrong master record
• Increased Confidence: With our detailed preview, you'll feel much more confident about merging decisions
• Time Savings: Our new intuitive interface means you'll spend less time reviewing and confirming merge operations
• Greater Control: You can now remove items from duplicate groups whenever you want, giving you more flexibility with your data

All these improvements make it way easier to keep your data clean and accurate, with much less effort on your part.

New Features and Improvements

New Features

Well, no new features, but 2 new sources entered Datablist this month.

Let’s begin with our new data sources and what they’re good for:

LinkedIn Jobs Scraper

Now you can scrape LinkedIn jobs as a Datablist source.

How to Use It:

1. Create a new Collection
2. Click on “See all sources”
3. Choose “LinkedIn Jobs Scraper”

Why we did it: We thought it just made sense, after having an Indeed source – the biggest job board worldwide – already available as a source, having also a LinkedIn Jobs source – the second largest.

Remote CSV/JSON source

Now you can connect remote CSV/JSON sources to Datablist and keep your collections synchronized with external data sources.

This feature is particularly valuable for teams working with:

• Multiple data sources across different platforms
• Frequently updated datasets
• Automated reporting systems

How To Use It:

• Create a collection
• Click on “See all sources"
• Select "Remote CSV/JSON Import"

Send me feedback ⇒ ⇒ Habib’s LinkedIn