Exploring what’s next

Innovation Hub

We explore, validate and turn emerging technologies into solutions with real-world impact.

How we innovate

binoculars

Explore

We identify and assess emerging technologies and market trends.

flask

Experiment

We build prototypes and proofs of concept in controlled environments.

checks

Validate

We assess technical feasibility and measure the potential impact of each solution.

circuitry

Industrialise

We turn proven outcomes into accelerators and solutions ready for our clients.

rApps

Telecom Lab

Intelligence for the networks of tomorrow

Telecom Lab is where we explore the evolution of telecom networks and how AI can make them more autonomous, efficient and sustainable.

We work across cloud-native architectures, Open RAN, network automation, SMO platforms, rApps, advanced data analytics and AI-assisted operations, collaborating with leading vendors and operators across the industry.

  • AI for network operations.
  • Open RAN and O-RAN ecosystems.
  • rApps and intelligent automation.
  • Network analytics.
  • Energy optimisation.
  • Private 5G networks.
  • Edge computing.
  • Zero-touch operations.

Projects

Radio Resource management in O-RAN

View project →
Virtualised resource management in O-RAN

View project →
Integrated E2E network-wide management

View project →
Energy-efficiency and security in O-RAN

View project →
head-circuit

AI Lab

Turning AI potential into business value

AI Lab is where we design, train and validate AI-powered solutions before bringing them into real-world projects.

We explore emerging architectures, foundation models, AI agents, multimodal systems and Responsible AI platforms to build reliable, scalable solutions aligned with our clients’ objectives.

Our approach combines applied research with a strong business focus, helping organisations accelerate AI adoption while maintaining security, control and measurable outcomes.

  • Generative AI.
  • AI agents.
  • RAG and knowledge systems.
  • Intelligent automation.
  • Multimodal AI.
  • Computer vision.
  • MLOps and LLMOps.
  • Responsible AI and AI governance.

Projects

Speech-to-Text REST API

A project using speech-to-text (STT) technologies, including automatic speech recognition (ASR) and speaker diarisation, to transcribe court proceedings and identify individual speakers.

A REST API was developed to enable straightforward integration with existing systems.
AIRE – AI Router Expert

An AI-powered chatbot built using RAG technology that acts as a ‘Router Expert’, helping end users configure their routers across the models supported by the company.

The solution implements a Retrieval-Augmented Generation (RAG) architecture based on an open-source LLM (Llama 3.1 70B), a Qdrant vector database and a Python-based controller exposed through a REST API.
R&D and Innovation Project: SIEM+AI

The solution detects and classifies cyberattacks within network traffic. Neural networks are applied to both public and synthetically generated datasets to identify network traffic anomalies.

The relevant features used for binary classification are surfaced for expert validation.

Detected attacks are then clustered using unsupervised learning techniques and large language models (LLMs) to determine the specific type of attack.
broadcast

O-RAN Lab

A real network for experimentation

Through our Telecom Lab, we have built our own O-RAN infrastructure to research, develop and validate new network capabilities in a controlled environment.

Our lab replicates the key components of an Open RAN architecture, enabling us to work on automation, network intelligence, resource optimisation, and new xApps and rApps before deploying them in operational environments.

  • O-CLOUD: Telco-optimized K8s Cluster.
  • SMO/OAM: O-RAN SMO.
  • RIC: O-RAN near-RT + Non-RT RIC.
  • Service Models: KPM (KPIs Measurement), RC (Radio Control) and CCC (Cell Configuration & Control).
  • xApps: Monitoring, slicing, LB, Traffic and QoS.
  • rApps: FlexPolicy, FlexMon.
  • gNodeB (vBS): OAI (Open Air Interface), SRSran.
  • RRU: USRP B210.
  • Mobile phones, commercial CPE and Quectel 5G module.
  • vRAN: shared computing platform composed of a commercial off-the-shelf (COTS) computer equipped with a 16-core Intel i9 14900K CPU.
  • near-RT RIC: Co-located with the vBS and implemented using FlexRIC.
  • SMO: hosts the non-RT RIC. For the non-RT RIC we use the open source implementation from the O-RAN Software Community (OSC), while the 5G Core VNFs can be deployed using Open5GS or OAI as Docker containers.
  • Testbed: includes an application server that can be used for hosting applications, traffic generation, and data storage.

Let’s talk

We look at every challenge in its real-world context to determine what’s worth building — and the right way to build it.

Tell us about your project