Enriching existing and new applications with AI integrations

The soundest way to get business value from AI is to place it inside the applications teams already use, rather than as a separate product. Tulpar adds AI-powered features to existing or new web and mobile applications that speed up decisions, cut repetitive work and always leave the final decision to an authorised user.

Reference application: AI-assisted ticket assignment in Tattoonets

In Tattoonets, built by Tulpar, the AI weighs each incoming ticket’s type, style preference, the artist’s expertise, calendar load and the customer’s previous artist, and suggests the best artist. The manager can approve or change the suggestion, and manual corrections feed into later suggestions.

AI features we can add to your applications

• Recommendations and automatic assignment: suggesting the best person or team for requests, tickets or tasks based on expertise, workload and history.
• Classification and prioritisation: sorting incoming requests, complaints and messages by topic and urgency.
• Document and content processing: extracting structured data from documents, forms and free text; producing summaries and draft content.
• In-app assistants: helpers that let users ask questions about application data in natural language, bound by permission rules.
• Reporting support: summarising trends and anomalies in operational data for managers.

Integrating ChatGPT, Gemini and similar models

We connect large language models such as ChatGPT, Gemini and Claude to your application through their APIs. The model only reaches data the user is authorised to see, follows your application’s roles and permissions, and presents its answers inside your own screens. We choose the model together, based on the use case, data privacy requirements and cost.

Business chatbots and assistants

We build chatbots that give your customers instant answers to common questions, and assistants that let your teams ask questions about application data in natural language. When the assistant cannot answer, or a decision is needed, it hands the request to an authorised person.

Process automation with AI

We automate repetitive work such as classifying incoming e-mails and requests, extracting data from documents, assigning records to the right person and preparing report summaries, while keeping approval steps in place.

Our principles

• Human approval: for critical decisions the AI suggests; approval and changes stay with an authorised user.
• Data security and privacy: we define at the start which data is processed for which purpose, and handle personal data within KVKK obligations.
• Integration with existing systems: we connect the AI layer to your application’s roles and permissions, database and APIs.
• Measurability: for each feature we agree indicators to track, such as suggestion acceptance rate and task time.

Approach

We first identify, together with you, the workflow where AI saves the most time and the data that feeds it. We then run a pilot with a limited scope and extend the feature based on user feedback.

Related guide: What is AI? How businesses add AI to their applications