Skip to main


The recruitment landscape is rapidly evolving, and modern companies require technology that can adapt to these changes to maintain a competitive edge in the hiring and talent management process. While resume parsing has long been a staple, Textkernel has been at the forefront of global-leading solutions in this area. Yet, the contemporary demand for more nuanced information about candidates has emerged as the new frontier in recruitment competitiveness.

Large Language Models (LLMs), like ChatGPT, offer a unique blend of human-like domain understanding and industrial scalability that allows parsing to be taken to a new level. This advancement allows for a wealth of information to be extracted from resumes, leading to a deeper understanding of candidates.

Introducing Textkernel's LLM Parser (beta)

Welcome to the future of resume parsing with Textkernel's LLM Parser (beta), where we seamlessly fuse Textkernel's robust, enriched proprietary parser with the capabilities of ChatGPT. This revolutionary fusion brings you a parsing solution like no other.

With Textkernel's LLM Parser, we've taken the power of Textkernel's Deep Learning expertise, proprietary taxonomies and two decades of parsing expertise, and married it with ChatGPT's natural language understanding prowess. The result is a parsing solution that offers unparalleled accuracy, potential limitless flexibility in data extraction, and the ability to unlock a world of possibilities in talent acquisition.

Explore the future of parsing technology with Textkernel's LLM Parser and experience the synergy of expertise, innovation, and unmatched performance. Get ready to revolutionize the way you parse resumes and embark on a journey towards better, more efficient recruitment. Your next-generation parsing solution awaits.

"We carefully crafted a parser that combines the strengths of LLMs with our industry-leading parsing technology and skills and professions intelligence. The result is a better parser than can be made with either technology independently."
- Ruud Liebregts
Chief Product Officer
Ruud Liebregts, Chief Product Officer
Start ParsingCTA Arrow

Bringing together the best of both worlds

Textkernel's LLM Parser (beta) represents the best of two exceptional worlds. On one hand, we offer our trusted, fast, dependable, and compliant talent data capture. On the other hand, we harness the global knowledge and precise natural language comprehension in LLM models like GPT.

We take care of the entire parsing pipeline, which entails more than just LLMs:

  1. Document conversion: Using industry-standard solutions we convert input documents to text (for parsing) and HTML (for display). Combined with Textkernel's Deep Learning engine that detects column layouts and document structure, we ensure that the LLM engine receives well-formatted text as input.
  2. Parsing: Using fine-tuned prompts optimized for accuracy and speed, we parse key data points using GPT-3.5.
  3. Normalization: the extracted data undergoes normalization through Textkernel's modules, which ensures consistent output. The data is enriched with mappings to taxonomies related to job titles, skills, education levels. While we also map to standards like O*NET and ISCO, Textkernel's real-world taxonomy of skills and job titles sets us apart.
  4. Inference: Our algorithms accurately compute additional metadata, such as the most recent use of a skill or the total months of experience with a skill.
  5. Output: The result from Textkernel's LLM Parser maintains the same JSON structure as our standard parser and seamlessly integrates with other Textkernel products.
I'm ready to start parsingCTA Arrow

Uncover a new world of Parsing possibilities

Textkernel's LLM Parser (beta) paves the way for future game-changing parsing developments. Imagine having precise answers to specific questions when reviewing a candidate profile:

  • Has this person ever worked in a Fortune 500 company?
  • Is this candidate experienced in B2B environments?
  • Has this applicant ever worked abroad?
  • Does a candidate have all certifications needed for regulated jobs like nursing?

Textkernel's LLM Parser (beta) is the foundation for future resume parsers that will be just as precise as they are flexible and customizable. Join us on this journey.

Why Textkernel and Textkernel's LLM Parser?

Continuous Innovation: At Textkernel, innovation fuels our growth and inspires us to continuously improve our solutions. We are the first established parsing provider to adopt LLMs, demonstrating our deep-rooted commitment to innovation for a competitive edge and allowing you to focus on your business.

Ease of Use: Textkernel now offers multiple world-class parsing solutions, each meeting specific needs but all delivering superior results. Benefit from quick, self-provisioning setup from the leading parsing provider, delivering results optimized for speed and accuracy.

Future-Readiness: Benefit from future proof technology without complicated API setup, or prompt engineering and LLM selection investments. If you require specific and tailored information parsed from your CVs and applications, Textkernel's LLM Parser provides the foundation for answering these questions.

Trust and Expertise: Textkernel's LLM Parser is built upon the solid foundation of our specialized resume parsing technology. Over the past two decades, we've consistently been at the forefront of adopting AI advancements, including Deep Learning, and introducing them to the market responsibly and fully compliant. In this context, the LLM Parser represents a continuation of our ongoing commitment to innovation and excellence.

Experience the possibilitiesCTA Arrow

Next level CV Parsing Excellence with a trusted foundation

As the global leader Resume Parsing solutions, we're trusted by 8 out of the top 10 global staffing agencies, parsing over 2 billion resumes each year. With the emergence of LLM technologies, we're excited to offer our customers a fusion of the best of both worlds, marrying our world-class parsing technology with the incredible world knowledge and language processing of LLMs, delivering an unparalleled experience for our 2.500+ customers.

How to get access to Textkernel's LLM Parser?

Create a demo account on the Sovren platform and you are ready to experience our latest parsing innovation. Connect with us to discuss how Textkernel's LLM Parser can help you achieve your hiring goals!

Frequently Asked Questions And Use Cases

Textkernel's LLM Parser demonstrates the advanced capabilities of our technology. You can assess whether it's suitable for your use cases or if our standard parser remains the better choice.

Our standard parser is known for its speed and impressive accuracy, achieving over 95% accuracy for the most critical data points. Textkernel's LLM Parser elevates accuracy even further, reducing the remaining errors by up to 30%. However, it's important to note that it currently requires more time for parsing and comes at a higher cost.
If you're already using the Sovren API, there's no need for modifications. The LLM engine utilizes the same API models and endpoints as the existing Sovren API. You can enable Textkernel's LLM Parser with a single boolean flag on the request.
Yes, the CV/Resume is sent to a private OpenAI Azure model and is not stored. OpenAI does not have access to this data, and it is not used to improve their models.
Initially, our focus was on English to showcase the capabilities of LLM technology. However, we plan to expand the LLM engine to include more languages in the future.
No, when you enable the LLM Parser and we detect a non-English document, we will return an error and not charge the add-on transaction cost.
View all FAQsCTA Arrow
Ready to get started?

Sovren Accelerator Program

Parse, index, search, and enrich up to 5000 documents for $200
This is a great way to get started with Sovren. The accelerator program gives you full access to our parsing software and allows you to test our AI Matching and Data Enrichment APIs as well. This is a one-time purchase worth up to a $1000 value.
Group 23Unlimited matching transactions.*
Group 23Unlimited boolean text searching.*
Group 23Integration consulting services and support.
Group 23Full access to JAVA and .NET SDKs.
*All subject to the Acceptable Use Policy