Newsroom Archive

recognition

AlpacaX Selected for Didimdol R&D Program

We’ve been selected for the Didimdol R&D Program, South Korea’s leading early-stage tech initiative. This recognises our work in AI-powered infrastructure security through Alpacon.

11 July 2025


image-1

A Program That Funds Readiness, Not Just Ideas

Run by South Korea’s Ministry of SMEs and Startups, the Startup Growth Technology Development Program – Didimdol R&D – is known for its rigorous selection process and focus on real-world execution. This program is one of the most selective in the country, favouring teams with original technical approaches, detailed execution strategies, and demonstrable public value. Our proposal was selected under the DIPS 1000+ category, a specialised track designed to accelerate breakthroughs in South Korea’s future-priority technologies.

With this support, we aim to detect abnormal behaviour in privileged server access through a realistic, phased plan. While AI in security is both timely and essential, AI-driven proposals are often perceived as overly abstract – or worse, trend-chasing – unless grounded in concrete execution. That’s why validation matters, and why our selection is particularly meaningful.

Our proposal not only outlined what we aimed to build, but also detailed a technically sound approach for doing so – earning us funding of up to £100K over 18 months to make it happen.

Where Access Happens – and Risk Begins

In modern infrastructure, the console is one of the most powerful – and most vulnerable – access points. It is where developers and administrators interact directly with critical systems, often with elevated privileges. Yet despite its importance, console activity often remains under-monitored. Traditional systems focus on who accessed what, but often overlook what users actually do once inside. That gap leaves room for subtle misuse – and tangible exposure.

A Platform Built for Scale and Intelligence

image-2

We’re designing our SaaS-based architecture to address that gap directly. Unlike on-premise systems that limit what can be collected, our platform allows for real-time aggregation of command activity, session flows, and behavioural signals, which are centralised across diverse environments. This not only improves visibility today – it also lays the groundwork for AI-powered abnormal behaviour detection tomorrow. Because our models aren’t tied to any single organisation, they can learn from usage patterns across organisations – at scale and in context.

From Vision to Execution

We’re building on a proven, in-production architecture and evolving it into a smarter, policy-aware session layer. Our roadmap includes abnormal behaviour detection, dynamic policy application, and developer-friendly feedback mechanisms. Each component will be validated through real-world PoCs and tested across diverse infrastructure environments.

With Didimdol R&D support, we now have the opportunity to put this plan into action – backed by government support, real-world validation, and public confidence. It’s a meaningful opportunity to demonstrate that infrastructure security can become more intelligent and adaptive, without compromising stability or trust.