T Challenge 2024: Pioneering AI Ideas for Telecommunications

Deutsche Telekom and T-Mobile US proudly announce the final twelve nominees for the T Challenge 2024 on AI for Telecommunications. The competition has seen an overwhelming response this year, with an influx of innovative ideas from universities, research institutes, spin-offs and a wide range of companies, showcasing a broad spectrum of groundbreaking ideas.
This esteemed competition presents a unique opportunity for teams worldwide to present their groundbreaking AI solutions for telecommunications. The nominated ideas, which are set to reimagine telecom networks and services, will be competing for the prestigious award for the best use of AI to augment telecommunications.
The second phase of the competition will see the finalists fine-tuning their pioneering ideas until June and culminates in a live exhibition and stage presentation at Deutsche Telekom’s headquarters in Bonn on June 10 and 11, 2024.
An esteemed panel of judges will recognize and honor the best ideas across various award categories. In addition to cash prizes totaling up to €225,000, all participating teams will have the chance to explore further business opportunities with Deutsche Telekom and T-Mobile US – across Germany, Europe, and the U.S.
T Challenge serves as a powerful symbol of the incredible potential and diversity of AI in the telecommunications industry. It’s a rare chance for innovators to showcase their advancements on an international platform and contribute to the future of the industry.
Join us for the demo day and award ceremony, where we’ll be celebrating the innovative spirit and groundbreaking solutions that have the potential to redefine telecommunications as we know it.

T Challenge Nominees 2024

Anymate Me


Anymate Bot - The first eco-friendly video chatbot in real time

Anymate Bot offers real-time video production for chatbots without a powerful GPU. Its patented ML method promises efficiency and higher FPS than Nvidia’s model. Promoting inclusive communication, the bot creates AI videos in 65 languages, including sign language. The USP is a patented technology, low power consumption, and multilingual capabilities.


United Kingdom

AI Continuum for Smarter Networks

Network representation and modelling are key to building efficient AI pipelines. A data-centric approach allows CSPs to optimize scenarios without duplicated pipelines. Our solution restructures network management through network graph modeling, network knowledge graph, graph discovery, and network configuration with Generative AI.



GNSS & 5G hybrid smartphone Precise Positioning

New smartphones can access real-time GNSS data and perform RTK positioning. By integrating this with the smartphone’s IMU data and the 5G NR data, GeoSPS can achieve a meter level accuracy in positioning. Through the application of AI techniques for integrating GNSS and 5G data, positioning accuracy can be improved to decimeter level outdoors and meter level indoors.



Computer vision AI for enhancing telco networks and field work

Inveniam uses computer vision AI for efficient asset monitoring and field work quality assurance. The technology integrates into workflows, offering real-time operation assessment and automated oversight.

Katonic AI


Multimodal chatbots using Katonic AI

Katonic AI’s multimodal function enhances chatbot services, making them more accessible for diverse customer needs. It can analyze objects and scenes within customer-uploaded media for self-serve troubleshooting, serve relevant guides, share video tutorials, and write descriptions of media for incident reports and data analysis.

NLPearl AI


A truly human-sounding, autonomous, zero latency conversational AI agent for phone sales and support

Enterprises facing high call volumes can leverage NLPearl AI’s agent, Pearl, for optimized sales and support. Pearl uses advanced AI and NLP to provide instant responses, reducing wait times and enhancing customer experience. Seamlessly integrated, this AI-powered call center solution is agile, cost-efficient, and offers significant ROI.

Rockfish Data

United States

Unlocking the Potential of AI-Driven Network Management With Synthetic Data

Rockfish Data’s AI platform uses synthetic data to improve network performance and security, enabling cost reductions and new revenue streams. The platform aids network operators in faster AI workflows, better incident management, and improved infrastructure security. It also ensures privacy and flexibility in data generation.


United States

Generative AI for Timeseries and Network Log Data - Search, Synthesize a Digital Twin, and Forecast

Synthefy, a Stanford/UT Austin spinout, offers a multi-modal foundation model for time series, enabling search, forecasting, and synthesis of privacy-preserving time series from a simple text prompt. It boosts accuracy by incorporating rich contextual metadata that today’s tools can’t handle. Applicable to any target group, it only requires a sample of past data.

Tiami Networks

United States

High Dimension Diffusion for Cross Network Evaluation

High Dimension Diffusion offers improved resolution for time-series data compared to traditional methods, as seen in studies like Hatanaka et al. Applied to high-dimension 5G data from core-network using 5G Core NFs and OAM systems, it can significantly enhance NWDAF by predicting anomalies ahead of time or just-in-time.



Building TRUST: a digital twin for AI validation in Network management

TRUST is a digital twin tool simulating smart city behavior across various users and services. It uses ML for dynamic scenario simulation and adaptation. Operating on global datasets, it generates synthetic data for city simulation, validating AI in network management simulations.



AI Care

Tupl’s AI Care provides real-time root cause analysis of customer issues, increasing first contact resolution by 5% and reducing level 2 escalations by 90%. With a response rate 100x faster than humans, it achieves 80% automation and proactive problem detection. R&D focuses on improving interactivity and new user journeys.

NEWT Lab, University of Washington

United States

RayTrack: Few Shot Data Augmentation for Multipath-Rich Wireless Environments

Imagine predicting wireless metrics in diverse scenarios using innovative techniques. The challenge? Sparse wireless measurements. RayTrack augments virtual data that mirrors real-world scenarios, enhancing services like 5G deployment, location tracking and wireless communication.

More News

T Challenge 2024: Grand Jury Announcement

On June 10-11, 2024 Deutsche Telekom headquarters in Bonn, Germany, will be the stage for twelve finalist teams. They will present their innovative solutions on ...
Read More

T Challenge 2024 seeks AI solutions for network management and customer experience

“T Challenge” by T-Mobile US and Telekom focuses on applications with Artificial Intelligence
Read More

T Mobile US and Deutsche Telekom Reveal 2022/23 T Challenge Winners

Six trailblazing companies driving Web3 innovation with the power of 5G receive top honors at the third annual global competition.
Read More