Leobit
AI-Powered CV Converter
Custom development of Leobit's internal tool for CV reformatting
ABOUT the project
- Client:
- Leobit's Internal Project
- Location:
-
USA
- Company Size:
- 100+ Employees
- Industry:
-
Information Technology
- Solution:
- Custom Software
A PoC for an AI-powered assistant for transforming CVs into Leobit’s corporate template. AI algorithms process a CV file and automatically restructure it into a project-based format while anonymizing contact information and applying corporate styling.
The solution enabled us to explore how AI models can organize diversely structured data, while the ReAct approach helped enhance output quality. The PoC demonstrated strong efficiency and precision, and we expect to reuse the underlying framework in future projects.
Customer
It was Leobit’s internal project that would facilitate our HR department in managing employee and candidate CVs. It also helped us build a framework that can be used for building custom solutions for our clients.
Business Challenge
Retrieving information from employee or candidate CVs and organizing it in the corporate format is often a time-consuming process. Leobit’s R&D team came up with a vision of a solution that would automate these processes while ensuring the formatting consistency of CVs.
Project
in detail
Our team has developed the solution’s PoC in around 2 weeks. We can roughly divide the work on the project into three stages.
Our R&D examined the needs of our project managers and HR department to understand the key issues with collecting and organizing data in software engineers’ and candidate CVs. We defined the solution’s architecture and chose the main technologies for its implementation, namely Angular and Tailwind for front-end development and .NET for back-end development. OpenAI GPT-4o model and the Document Intelligence feature of Azure Functions were chosen for AI-powered document processing and reformatting.
Upon planning the solution, we started active development. Our specialists built the app’s front end and back end and integrated it with AI models. We also used the Spire.Doc library for creating DOCX documents as an output. Our team applied the ReAct (Reasoning and Acting) approach to enable iterative validation of generated content, improving the quality of the tool’s outputs.
Once the solution’s core was developed, our specialists ran several tests and optimized the tool to ensure better performance and output accuracy. In addition, our team thoroughly documented all the stages of the development process. In the future, we will be able to use this documentation to recreate the project framework in other development projects oriented towards our clients.
AI-POWERED DOCUMENT PROCESSING
The tool uses the capabilities of the OpenAI GPT-4o model and the Document Intelligence feature of Azure Functions to read and understand CVs in different formats. It parses content, identifies key sections, and retrieves the information from them automatically, saving time and reducing manual effort.
ANONYMIZATION OF CONTACT INFORMATION
All personal and sensitive data, such as names, emails, and phone numbers, is automatically anonymized. This ensures privacy and allows CVs to be safely shared or stored internally without exposing personal details. This feature is especially relevant if the framework used for building the PoC will be reused for some large freelance platforms or fintech solutions dealing with strict privacy regulations.
CORPORATE STYLING AND PROJECT-BASED FORMATTING
The system applies Leobit’s corporate template, including layout, fonts, colors, and branding. CVs are converted into a professional, project-focused format that highlights relevant experience. The solution ensures consistency of reformatted CVs. Meanwhile, formatting settings can be easily optimized or adjusted to adapt to changing project requirements.
Explore The solution Prototype
Custom development of a PoC for an AI-powered solution that reformats CVs into a corporate template, anonymizes contact information, and standardizes candidate profiles using document intelligence and LLM reasoning.
Technology Solutions
- Convenient UI built with Angular and Tailwind CSS.
- CV processing and reformatting powered with the OpenAI GPT-4o model and the Document Intelligence feature of Azure Functions.
- The Spire.Doc library that automatically creates PDF documents as an output.
- Support for files in PDF and DOCX formats.
- Iterative validation of generated content with the ReAct (Reasoning and Acting) approach.
Value Delivered
- An efficient and precise tool that supports internal HR workflows.
- A framework that can be reused for building solutions across domains, in particular, in consulting, education, financial services and banking, or freelance marketplace development.
- Strong focus on privacy regulations with anonymization capabilities.
- Ability to expand the tool’s functionality to process, summarize, and restructure other document formats.
