Knowledgator Engineering

Name

Knowledgator Engineering

Legal name

Knowledgator Engineering

Website

https://knowledgator.com

Company type

Private

Founded year

2023

Operating status

Operating

Headcount range

1–10 employees

Ownership category

akta.pro rank

Product category

Natural Language Processing / AI Information Extraction

NAICS

Software Publishers (513210)

SIC

Services-Computer Programming, Data Processing, Etc. (7370)

akta.pro primary industry

Text Analytics & Information Extraction (NER, classification, sentiment) (HDAAADAA)

akta.pro secondary industries

OCR, ICR & Intelligent Document Processing (IDP) (BPAAAKAB), Search, Retrieval & Semantic Ranking (BM25/vector, hybrid) (HDAAADAB), Data Discovery, Classification & Labeling (HDADAFAF), Bioinformatics & Multi-omics Analysis Software (HLAGAJAD)

Keywords

• Named Entity Recognition
• Information Extraction
• Zero-Shot Text Classification
• Open-Source NLP Models
• Knowledge Graph Retrieval

Headquarters

HQ city

London

HQ country

United Kingdom

HQ region

Europe

Markets served

GTM type

B2B

Offering type

Software

Cost components

Technology or R&D, Personnel, Infrastructure, Marketing or Sales, Operations

Revenue model

1. Platform Subscription/Tiered Plans: SaaS platform with multiple subscription tiers ranging from individual developers to enterprise-scale deployments. Free credits provided for initial exploration. Customized plans available for high-volume or specialized requirements.
2. Usage-Based API Access: Pay-per-use model for API access, with inference credits consumed based on volume. Supports pay-as-you-go and monthly billing options.
3. Enterprise Consulting Services: Consultation with team for unique use cases, custom integrations, and specialized deployment requirements.
4. Open-Source Models (Freemium): Core open-source models available for free download and local deployment, with paid cloud platform access for those preferring managed infrastructure.

Pricing tiers

Go-to-market motion2 records

Distribution channels4 records

Marketing channels10 records

Core offering

Knowledgator Engineering builds and distributes open-source AI/NLP models and a commercial SaaS platform that extract structured data (entities, relationships, classifications, JSON) from unstructured text using a non-generative approach with sub-second latency. The portfolio centers on the GLiNER zero-shot Named Entity Recognition family, the GLiClass zero-shot text classification family, Extract-It relation extraction, Text2Json structured output conversion, RetriCo knowledge graph retrieval, and specialized variants (multilingual, biomedical, PII detection). The Knowledgator Platform wraps these into a managed service offering fine-tuning, hardware optimization, and one-click deployment for enterprise buyers.

Product overview

Knowledgator Engineering offers an open-source AI platform for information extraction, comprising multiple specialized products that work together as both open-source frameworks and a commercial platform. At its core, the platform is powered by GLiNER (zero-shot Named Entity Recognition) and GLiClass (zero-shot text classification), supplemented by Extract-It (relation extraction), Text2Json (text-to-structured-JSON), GLinker (entity linking for knowledge base connections), RetriCo (knowledge graph framework for building and querying RAG systems), LiqFit (few-shot fine-tuning), and Chemical Converter (chemical format translation). The commercial Knowledgator Platform wraps these open-source models into a unified SaaS offering, providing API access, fine-tuning, hardware optimization (quantization/pruning/distillation), and one-click deployment across on-premises, private cloud, and edge environments — all delivering structured JSON output at sub-second latency. Specialized variants target specific verticals: GLiNER-X (multilingual, 20+ languages), GLiNER-PII (privacy/PII detection), and GLiNER-BioMed (biomedical domain, developed with University of Geneva). The platform collectively reports over 5 million model downloads on Hugging Face.

Differentiator

Problem solved

Functional benefit

Brands

• GLiNER: Efficient zero-shot Named Entity Recognition (NER) models for extracting entities from text.
• GLiClass
• GLiNKER
• LiqFit
• RetriCo
• Chemical Converter

Products and services

• Knowledgator Platform All-in-one commercial SaaS for optimizing AI models for edge and on-premise deployment: fine-tune open-source models, optimize them for any hardware (quantization, pruning, distillation), and ship to on-prem, private cloud, or edge environments through a single interface with API access and multiple pricing tiers.
• GLiNER Open-source zero-shot Named Entity Recognition model family that detects and classifies entities such as people, organizations, locations, dates, monetary values, and custom user-defined types across multiple languages, with specialized variants (bi-encoder, multilingual, biomedical, PII).
• GLiClass Open-source zero-shot text classification model family that categorizes documents, paragraphs, or sentences into any label set without fine-tuning, supporting multi-label classification, topic categorization, sentiment analysis, content moderation, and customer-intent classification.
• Extract-It Relation extraction model that maps how entities are connected (e.g., which gene regulates another gene) and automatically constructs knowledge graphs, integrated into the unified Knowledgator pipeline alongside NER and classification.
• Text2Json Text-to-structured-JSON conversion capability that transforms free-text input (invoices, contracts, clinical notes, financial filings) into clean, schema-conforming JSON suitable for databases, APIs, and downstream enterprise systems.
• GLinker Open-source entity linking framework with a layered pipeline (L1 NER, L2 candidate retrieval via dictionary exact/fuzzy matching, L3 disambiguation via linker model, L0 aggregation) that connects extracted entities to knowledge bases, hospital systems, medication formularies, and terminology catalogs.
• RetriCo Open-source knowledge graph framework for building, querying, and managing knowledge graphs, with a tool-calling layer that lets LLM agents issue structured function calls translated into parameterized Cypher queries, and supports graph database backends, community detection, and KG embeddings.
• Chemical Converter Specialized open-source model collection that translates between different chemical naming and structure formats (e.g., chemical name ↔ standard chemistry interchange formats) for life-sciences and chemistry workflows.
• LiqFit Open-source few-shot learning framework for fine-tuning Knowledgator models with minimal labeled data, achieving results comparable to supervised models that require 10–100x more training examples.
• GLiNER-X (Multilingual) Multilingual Named Entity Recognition model built on MT5 encoder architecture supporting 20+ languages (Swedish, Norwegian, Czech, Polish, Lithuanian, Estonian, Latvian, Spanish, Finnish, English, German, French, Romanian, Italian, Portuguese, Dutch, Ukrainian, Hindi, Chinese, Arabic) in three model sizes.
• GLiNER-PII Production-grade model for detecting personally identifiable information (PII), protected health information (PHI), and payment card industry (PCI) data across 60+ predefined categories, supporting HIPAA, PCI-DSS, and GDPR compliance workflows.
• GLiNER-BioMed Biomedical Named Entity Recognition model family developed in collaboration with the DS4DH research group at the University of Geneva, leveraging synthetic annotations distilled from large generative biomedical language models to deliver zero-shot and few-shot biomedical entity recognition in six model sizes.
• GLiNER Multi-task Large Model v0.5 Multi-task NER model specialized for invoice data extraction, capturing invoice numbers, dates, due dates, vendor and customer names/addresses, line items (description, quantity, unit price, line total), subtotal, tax rate, tax amount, and total, with export to QuickBooks, Xero, and Sage accounting systems.
• knowledgator-pipeline Core workflow engine of the Knowledgator Platform orchestrating a four-step pipeline: select a task-specific model (GLiNER, GLiClass, Encoders), fine-tune on domain data with minimal examples, optimize via quantization/pruning/distillation, and deploy to on-prem, private cloud, or edge with one click at sub-50ms inference.

Quantifiable outcome

• 98% F1 score on PII detection in Electronic Health Records vs 22.3% for Llama and 50.2% for Azure Presidio
• +4 more outcomes

Segments5 records

Ideal customer profiles2 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration8 records

AI capability5 records

Feature13 records

Partnerships

One partnership is on record.

• DS4DH (University of Geneva)coreStrategic or Co-development PartnerCollaboration with DS4DH research group from University of Geneva for developing GLiNER-BioMed specialized biomedical NER models. The partnership leverages synthetic annotations distilled from large generative biomedical language models to achieve state-of-the-art zero-shot and few-shot performance in biomedical entity recognition tasks.

Scale indicators7 records

Recent moves6 records

Expansion highlights6 records

Direct peers

• Hugging Face: Hugging Face is the primary open-source hub for transformer models including NER, classification, and information extraction. Knowledgator publishes its models on Hugging Face and competes for the same developer mindshare in open-source NLP.
• John Snow Labs: John Snow Labs provides enterprise-grade biomedical and clinical NLP models with HIPAA-grade deployment, directly competing with GLiNER-BioMed in the healthcare vertical.
• spaCy: spaCy is the leading open-source industrial-strength NLP library for NER, classification, and information extraction. It is a foundational alternative to Knowledgator's GLiNER/GLiClass for developers building extraction pipelines.

Broad incumbents

• Azure AI Language (Microsoft): Azure AI Language is a hyperscaler NLP service offering NER, PII detection, sentiment, and text analytics. It is a named competitor in Knowledgator's PII benchmarks (50.2% F1) and competes for enterprise extraction workloads.
• Amazon Comprehend: Amazon Comprehend is AWS's managed NLP service covering entity recognition, PII detection, and document classification. It is a broad incumbent competing for the same enterprise extraction workloads as Knowledgator's commercial platform.
• Google Cloud Natural Language API: Google Cloud's Natural Language API provides entity analysis, classification, and custom NLP within the broader Google Cloud ecosystem. It is a broad incumbent alternative, particularly relevant given Knowledgator's undisclosed Google investment.

Emerging players

• Cohere: Cohere provides enterprise NLP APIs for classification, NER, and retrieval. It targets similar enterprise extraction workloads with a managed cloud model rather than Knowledgator's open-source / on-prem approach.
• Snorkel AI: Snorkel AI focuses on data-centric AI with weak supervision and programmatic labeling for enterprise NLP. It overlaps with Knowledgator's few-shot/LiqFit positioning and its target buyer of enterprise data teams.
• Primer.ai: Primer.ai provides enterprise NLP and knowledge extraction for government, finance, and intelligence customers. It is a comparable specialist in information extraction with a strong on-prem / air-gapped deployment posture.
• Scale AI: Scale AI is a data-centric AI company with adjacent capabilities in labeling, evaluation, and fine-tuning. It overlaps with Knowledgator's LiqFit few-shot fine-tuning and data-advantage positioning.

Market position

Strengths4 records

Weaknesses4 records

Competitive moat5 records

Key risks6 records

Key highlights6 records

Customer concentration

• Xx.com/knowledgator
• GitHubgithub.com/Knowledgator
• LinkedInlinkedin.com/company/knowledgator

Compliance3 records

Revenue estimate

Valuation estimate

Number of profiles

Profiles3 records

Funding overview

Funding rounds5 records

Investors3 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Knowledgator Engineering