FPT Launches Flezi Foundry™, Advancing AI-Augmented Delivery Platform for Global Enterprises
Following its AI-first strategy, FPT has introduced Flezi Foundry™, an AI-agent-driven platform that handles tasks from software development to IT operations through a Service-as-a-Software model.
📋 Article Processing Timeline
- 📰 Published: May 28, 2026 at 03:46
- 🔍 Collected: May 27, 2026 at 19:15
- 🤖 AI Analyzed: May 27, 2026 at 19:15 (0 min after Collected)
FPT, a global IT service provider, has officially announced the launch of Flezi Foundry™ (also known as FPT Digital Foundry™), a new AI-augmented delivery platform. Built on a "Service-as-a-Software" model where software directly executes business tasks, the platform integrates autonomous AI agents with human expert supervision, secure infrastructure, and an outcome-based delivery mechanism. This initiative aims to modernize enterprise IT by positioning AI agents as key contributors to development and operations.
Flezi Foundry™ utilizes an "Agentic Engineering" approach to implement AI agents within software development and IT workflows. This structured methodology incorporates human oversight, governance, transparency, and performance measurement directly into the delivery process.
The platform is offered in two primary service modes:
1. Agentic Development Lifecycle (ADLC)
Under human supervision, specialized AI agents support the entire development lifecycle, including design, coding, review, testing, security, and documentation. This model aims to enhance quality and transparency, targeting a productivity increase of up to 30%.
2. Agentic Managed Services (AMS)
This mode implements a collaborative model where AI agents work alongside humans to handle IT operations. Agents assist in alert prioritization, incident response, recovery, and service improvement. Upon reaching maturity, the platform is expected to automate 60-90% of first-level support inquiries and achieve a 99.5% Service Level Agreement (SLA).
The platform's architecture rests on four pillars:
- A hybrid sovereign infrastructure combining Azure Virtual Private Cloud and FPT AI Factory.
- "Hybrid FTE Pods" where human supervisors manage AI agents through Human-in-the-Loop and Human-on-the-Loop models.
- A capacity-based pricing model linked to story point velocity and outcome-based SLAs.
- "Digital Brain" and "Skill Marketplace" for accumulating and reusing standardized procedures and expertise.
To minimize implementation risks, FPT provides a phased transition process beginning with discovery workshops, followed by pilot phases (A/B testing or shadow mode), before moving to full-scale deployment.
Frank Bignone, Senior Vice President and Head of Corporate Strategy & Growth at FPT Software, stated: "The emergence of Agentic AI marks a new era in enterprise IT services. Intelligence is now being embedded directly into operating models rather than just tools. Through Flezi Foundry™, FPT continues to build a future delivery model supported by coordinated governance between humans and AI agents."
This launch furthers FPT's growth strategy under its AI-first vision. By building on "FleziPT," its AI-first platform for enterprise transformation, FPT is extending AI application across the software development and IT operations domains.
Flezi Foundry™ utilizes an "Agentic Engineering" approach to implement AI agents within software development and IT workflows. This structured methodology incorporates human oversight, governance, transparency, and performance measurement directly into the delivery process.
The platform is offered in two primary service modes:
1. Agentic Development Lifecycle (ADLC)
Under human supervision, specialized AI agents support the entire development lifecycle, including design, coding, review, testing, security, and documentation. This model aims to enhance quality and transparency, targeting a productivity increase of up to 30%.
2. Agentic Managed Services (AMS)
This mode implements a collaborative model where AI agents work alongside humans to handle IT operations. Agents assist in alert prioritization, incident response, recovery, and service improvement. Upon reaching maturity, the platform is expected to automate 60-90% of first-level support inquiries and achieve a 99.5% Service Level Agreement (SLA).
The platform's architecture rests on four pillars:
- A hybrid sovereign infrastructure combining Azure Virtual Private Cloud and FPT AI Factory.
- "Hybrid FTE Pods" where human supervisors manage AI agents through Human-in-the-Loop and Human-on-the-Loop models.
- A capacity-based pricing model linked to story point velocity and outcome-based SLAs.
- "Digital Brain" and "Skill Marketplace" for accumulating and reusing standardized procedures and expertise.
To minimize implementation risks, FPT provides a phased transition process beginning with discovery workshops, followed by pilot phases (A/B testing or shadow mode), before moving to full-scale deployment.
Frank Bignone, Senior Vice President and Head of Corporate Strategy & Growth at FPT Software, stated: "The emergence of Agentic AI marks a new era in enterprise IT services. Intelligence is now being embedded directly into operating models rather than just tools. Through Flezi Foundry™, FPT continues to build a future delivery model supported by coordinated governance between humans and AI agents."
This launch furthers FPT's growth strategy under its AI-first vision. By building on "FleziPT," its AI-first platform for enterprise transformation, FPT is extending AI application across the software development and IT operations domains.
FAQ
What is the primary innovation behind Flezi Foundry™?
It utilizes a 'Service-as-a-Software' model where AI agents, managed by humans, directly perform software development and IT operational tasks rather than just acting as tools.
What efficiency gains can companies expect?
Development (ADLC) can see productivity improvements of up to 30%, while Managed Services (AMS) can automate 60-90% of initial support inquiries.
How does FPT ensure security and governance in this AI-led model?
The platform uses a hybrid sovereign infrastructure and 'Human-in-the-Loop' models to ensure every action taken by AI agents is supervised and transparent.