case study

AI-driven HR Tech company dramatically improves the quality of its recommendations by pulling data from multiple sources - iPaaS for SaaS

About is an embedded integration platform for SaaS companies. helps AI companies to add a wide range of integrations to their platform. These integrations can either capture data from multiple data sources (cloud applications) to feed the AI engine, or they can send results from the AI platform to other cloud applications. Since AI relies heavily on data, these integrations are a key element of the AI solution.

About the customer

The customer is a fast-growing AI company in the HR Tech space. The company applies AI on employee data to improve the productivity of Fortune 500 enterprises.

The integration challenge

The company was feeding their AI engine with data gathered from their own productivity tool. The company was looking for a solution to collect more data. The goal was to pull new data from other cloud applications as well, such as productivity tools, CRM and HR / ATS software etc. in order to provide more learning data to their AI engine.

To build all these integrations, the customer needed to engage its internal development team. “Classic” one-off custom development turned out to be very costly, hard to maintain, and impossible to customize for individual customers. Further more, limited knowledge on the various destination platforms, and no readily available test accounts significantly delayed the creation of these integrations.

martech integration

AI integration scenario and requirements

The AI company needed to extract large amounts of data from various cloud tools, including CRMs, HR and ATS software etc. The requirements included:

 Support for a large number of cloud applications

 Ability to customize an integration for a specific enterprise customer

Support for custom fields

Ability to add custom logic to the data flows

How the integrations were built on

The integrations were created and published using the platform in 4 steps.

Step 1: Training and support by

Step 2: Creation and testing done by the customer’s solution engineers

Step 3: Test environments provided by

Step 4: Integrations go live. Lead time: 4 to 6 weeks

Integration activation per end-customer


The SaaS company applies the “full-service delivery model” to activate integrations for individual customers. 

This means that the Customer Success Team of the AI Tech provider takes care of the activation of the integrations for each enterprise customers. A Customer Success Engineer goes through a user-friendly Setup Wizard in, to activate a new integration in 4 steps:

Who does what

Step 1: Send a secure “Invite” email to the end-customer, to request credentials to access the CRM, productivity tools and other platforms (e.g. Hubspot, Salesforce, MS Dynamics, Zendesk, etc.). Once the customer authorizes access, the new data sources are automatically connected.

Step 2: Apply specific “field mapping”: link custom fields in the AI platform to custom fields in the CRM.

Step 3: Choose the “matching” behavior in the AI platform and the CRM: e.g. match employees based on email, or a combination of name & email, etc.

Step 4: The integration is live and new data is constantly flowing into the AI engine.



Avoiding the “cold start” problem when onboarding new customers. The training data is pulled from existing data sources to feed the AI engine.

3-5 months lead time saved per customer by building integrations with instead of doing custom development.

 A dramatic increase in the quality of the outcome of the AI engine.

 The company now wishes to deploy a large number of connectors to various cloud applications.

 The AI tech provider closes more deals by delivering enterprise-grade integrations that prospects are asking for.