This page contains public marketing content of Pragatix
Pragatix Marketing Resources
Website: AGATSoftware.com
Video: View
Pragatix General Datasheet: View
Pragatix AI Firewall Datasheet: View
Pragatix Private AI Datasheet: View
General Presentation: View
Short pitch with screenshots presentation View
Wiki documentation: View
5 Minute Introduction Video and 15 Minute Business GPT Pitch
Try it yourself
Follow the guides below to try the product for free, including guided exercises:
Go to https://app.pragatix.ai
Wiki links
User guide - AI Firewall View
User guide Private AI View
The main components View
The deployment options (SaaS/ on-prem) View
AWS System requirements: View
On-prem system requirements View
Short Product descriptions
Slogan: AI Security & Enablement Platform
Sub-text: Private AI, Guardian Agent and Prompt Control, AI Productivity and Awareness
Pragatix AI Platform One line
Option 1
Pragatix is an AI Security & Enablement Platform that combines an AI Security Suite to secure and control AI usage and agents with a Private AI Suite to enable secure AI adoption while improving productivity and security awareness.
Option 2
Pragatix AI Platform enables organizations to safely leverage AI by providing a private, secure suite of AI services, backed by a firewall that provides full visibility and control over public AI usage.
Option 3
Pragatix AI Platform enables enterprises to adopt AI safely by providing private AI services and governance for public AI use, ensuring security, compliance, and operational control.
Private AI Suite one line
Enterprise AI services within the Pragatix platform provide organizations with ready-to-use AI capabilities with zero data exposure.
AI Security Suite one line
Provides visibility, control, and protection across AI agents, user interactions, and AI models, preventing risky actions, data exposure, and unsafe AI behavior in enterprise environments.
Short messaging
Empower your enterprise AI with Pragatix.
Ensure zero data exposure with the secure Private AI Suite while gaining complete visibility and control over public AI usage via the AI Security Suite.
AI Security Platform- slogans
Unified activity control across employees, AI agents, and models
Homegrown, SaaS, embedded, and endpoint coverage
Platform Overview: Paragraph
AGAT’s Pragatix is a security-first AI platform and a suite of enterprise AI services that enables organizations to adopt generative AI with full governance and data privacy control, supporting both on-premises deployment and controlled use of public AI services.
It combines two core solutions: an AI Firewall for secure use of public AI, and a Private/On-Prem AI Suite of AI services for organizations that require zero data exposure.
The AI Secuirty Suite provides three protection layers—Prompt Guardian, Guardian Agent, and Model Guardian.
The Prompt Guardian layer governs access to services like ChatGPT, Copilot, and Gemini in real time, the Guardian Agent layer enforces policy on AI agents and MCP-based tools, and the Model Guardian evaluates models through a risk engine and vulnerability assessments.
It enables risk-based policy enforcement, data classification, and visibility into enterprise and shadow AI usage, while protecting against threats such as prompt injection and data leakage.
The Private AI Suite is an end-to-end enterprise-ready AI platform deployable in air-gapped, private cloud, or SaaS environments. It connects to company data sources to power knowledge assistants, natural-language data analysis, smart search, AI agents, workflow automation, log and anomaly analysis, document processing and translation, code assistants, and automated data classification.
Platform Pitch for conference
AGAT offers Pragatix, which includes two main modules
An AI Security offering visibility and control over AI usage and agents, and a model, and separately for a more secure environment, a Private AI Suite that offers different AI services such as chats and agents that can be installed on-prem with no internet access, ensuring data privacy
We also offer tools for training users for security awareness improvements
Pragatix overview- Detailed
AGAT's Pragatix is a security-first AI platform that enables enterprise-generative AI services with on-premises deployment & public AI Governance.
Pragatix offers two types of solutions:
AI Security Suite – for organizations that want to use public AI while maintaining full control.
Private/On-prem AI Suite For companies that do not want to risk any data exposure.
AI Security Suite:
The Security Suite provides end-to-end protection across human interactions, AI agents, models, and data, covering SaaS, homegrown applications, endpoints, and developer IDEs.
The main capabilities include:
Prompt Guardian is a real-time governance for public AI services such as ChatGPT, Copilot, and Claude.
Guardian Agent governs AI agents and tool activities via an AI Gateway for real-time policy enforcement. Additionally, the Guardian agent provides discovery and enforcement of AI Agents using Network Proxy, Browser Extensions, API, and endpoint agents
Model Guardian includes a Risk Engine that evaluates AI models using source intelligence and static & Dynamic Red Teaming dynamic red teaming for vulnerability assessments.
AI Gateway enables developers and applications to securely access both internal and external AI services through a governed gateway that delivers full visibility, control, and cost and consumption management.
Security Awareness - Provides visibility into AI usage to detect risks and deliver targeted, real-time, in-context training directly within tools like ChatGPT.
The AI Security Suite enforces risk-based policies, classifies data sensitivity, and provides full visibility into both enterprise and shadow AI usage, while delivering built-in security controls against major OWASP AI threats, including prompt injection, data leakage, and model abuse.
Private AI Suite:
The AI Suite is an end-to-end, enterprise-ready local / on-prem AI platform
It ensures zero data exposure through flexible deployment options, ranging from fully air-gapped (no-internet) environments to private cloud and full SaaS deployments.
By connecting to a wide variety of data sources such as documents, websites, and databases, Pragatix’s Private AI delivers the following core capabilities:
Knowledge Assistant - Generate answers from all company-connected sources and pre-trained knowledge, preserving source permissions and grounding answers to your data.
Data analysis - Using natural language, perform complex data analysis on Excel and databases.
Smart search - Search all your content beyond Keyword Matching. Understand the intent and context.
AI Agents - Run AI agents to plan and perform tasks using Python code, such as internet search, image analysis, speech to text. Integrate with MCP servers and expose the platform as an MCP server.
Workflow builder - Build workflow using no-code drag and drop, calling AI Agents and other AI services
Log analysis- Ingest logs, classify and structure data, visualize, and apply real-time anomaly detection.
Data Extraction - Capture, classify, and extract information from unstructured or semi-structured documents for IDP- Intelligent Document Processing.
Document Translation- Translate documents preserving original layout and formatting
Anomaly Detection - Identify anomalies in data, such as logs
AI Code Assistant - Code completion, error detection, code generation, and answering questions from the company codebase
Data Classification - Automatically classify data based on existing data.
AI router- Smart route AI prompts to cloud and on-prem multi-model services
Behaviour intelligence (security & Adoption)
The Behavioral Intelligence module provides real-time visibility into how employees use AI tools across the organization, identifying risky activities such as sharing sensitive information, unsafe agent usage, or policy violations. By analyzing actual user behavior and prompt activity, the platform delivers measurable insights into AI risk exposure and usage patterns. When risky actions occur, employees receive immediate, in-context feedback and micro-training to help them correct their behavior in the moment. This approach transforms AI security from static policy enforcement into continuous learning—reducing risk while improving employee awareness and responsible AI usage.
AI Behavior Intelligence (Adoption Intelligence)
Pragatix AI Behavior Intelligence enables organizations to improve productivity by measuring and optimizing employee interactions with AI across the enterprise.
The platform delivers enterprise adoption dashboards with role-based visibility for managers and executives, tracking adoption trends across departments and over time. It provides deep insight into AI usage and user intent, automatically classifying how AI is applied in real business workflows and identifying high-value use cases that can be replicated and scaled.
With full AI service visibility, organizations can monitor which AI tools and platforms are being used—across both approved and public services—and understand cross-platform usage patterns.
Through adoption benchmarking and enablement, the platform identifies top-performing users and AI champions, as well as teams that require additional support. It delivers personalized insights, performance-based guidance, and continuous progress reporting to improve how employees use AI.
By combining measurement, visibility, and targeted enablement, organizations can continuously improve human-AI interactions—driving higher efficiency, better outcomes, and clear, measurable ROI from AI adoption.
identify power users who are getting the most out of generative AI features. These users could have generative AI tips, best practices, and workflows that can be shared with others in your organization.
Identify users with low adoption rates who might benefit from training and other materials related to generative AI.
About Pragatix by AGAT
Leveraging over a decade of compliance expertise, AGAT delivers the Pragatix AI Security & Enablement Platform, enabling secure, productive AI adoption with real-time control and AI sovereignty.
At its core, the AI Guardian Agent governs AI agents and user interactions, preventing risky actions and data exposure in real time.
Pragatix combines data privacy, agent control, and AI Security Awareness, providing visibility into AI usage, detecting risky behavior, and enabling in-context training to improve adoption safely.
Our mission is to empower enterprises with secure, responsible, and measurable AI usage, across both public and private AI environments.
AGAT Software is trusted by over 25 Fortune 500 companies worldwide.
About AGAT 140 characters
Leveraging a decade of compliance expertise, AGAT delivers the Pragatix AI Security & Enablement Platform, providing real-time control and AI sovereignty to enable safe, productive AI adoption.
About AGAT
AGAT is a leading provider of cutting-edge AI solutions. Its flagship product, Pragatix, is an enterprise-ready AI platform that enables rapid adoption of Generative AI while ensuring data privacy and security. Backed by more than a decade of expertise, AGAT has delivered trusted compliance and security solutions to dozens of highly regulated organizations, including over 25 Fortune 500 companies.
Firewall model risk engine details
Answer: The firewall runs dynamic and static security scans of the model.
This includes metadata scans of the model (license, stats, deep background research of the authors and of known security vulnerabilities etc.), static scans of the files (file types, scan for malware signatures, weight file formats used by the model), and dynamic analysis of the model (jailbreak tests, prompt injections, etc.).
Hebrew description