Tupl Inc.

Tupl's goal is to help operators increase their operational efficiency and reduce cost, leveraging their extensive knowledge of wireless technologies, designing and operating networks, and developing support tools. Toward that end, the company developed TuplOS, a scalable back-end network solution based on open architecture for access to data and increased process automation for network back-end systems. The company's software uses an AI engine, machine learning, and other utilities for the networks and their customers.

Tupl is headquartered in Bellevue, Washington and was founded in 2014 by Pablo Tapia, a telecommunications veteran who previously worked at T-Mobile. The company's mission is to be a leader in Intelligent Process Automation (IPA) for Telecommunication network operations. Since its founding, the company claims it has reduced manual labor by 90 percent, increased speed, and increased the accuracy of customer networks when compared to manual engineering processes.

Tupl's TuplOS is built to be a pragmatic approach to machine learning and is designed for domain experts to facilitate the creation of complex automation utilities within any industry. To achieve this, TuplOS is designed to be stable, scalable, and easy to maintain, offering easy data ingestion, simplified maintenance for minimum human intervention, and easy data processing. This is accomplished with a customizable user interface with integrated reporting and with artificial intelligence performing tasks that are not performed by data scientists.

TuplOS is built to leverage open-source big data components for storing and processing datasets distributed on large clusters of commodity hardware. This includes working with Docker, Kubernetes, Prometheus, Grafana, KeyCloak, Hadoop, Apache Spark, Apache Phoenix, Apache Kafka, Ceph, Apache Hbase, Apache Nifi, PostgreSQL, Rook, MongoDB, and Cockroach DB.

Tupl's AI Care is built to help network companies react to customers' issues before the customers can be seriously impacted. This includes automatic ticket resolution, including taking action for automatic problem identification, root causing, and routing of technical customer complaints to provide zero-touch responses to detected problems. It also offers recommendations to be used by technicians. As well, AI Care can automatically resolve customer complaints through monitoring suspicious patterns from system logs and creating tickets for reporting, analyzing, and tracking problems, to resolve issues before a customer notices.

Tupl's Power Saving Advisor, as the name suggests, works in networks to recommend areas or solutions to reduce power consumption for both open-loop and closed-loop operations and processes. The Power Saving Advisor works by generating automated tickets for every action needed for a cell or a site to reduce power consumption, using a pre-trained machine learning model that analyzes those areas to provide a solution for power consumption that maintains the customer experience.

Built for telecommunication networks, Tupl's Network Advisor offers automated actions from the company's artificial intelligence and machine learning models intended to diagnose and provide action plans to maintain and optimize telecommunication networks. Further, the advisor can be used to digitize processes for telecommunication networks while providing increased automation across a network with artificial intelligence.

Tupl's RF Shaping uses an artificial intelligence learning agent to reduce network interference while maintaining coverage areas and network consistency. It does this by reducing the levels of intra-cell interference in urban areas without degrading coverage, utilizing users' geolocalized information and network topology to achieve best results. The AI is trained with reinforcement learning to find improvement opportunities in a network.

Tupl's Unifier is built to track key data sources and provide an overview of those data sources in a single unified dashboard. The dashboard provides users with data correlations on multiple data sources from a network to provide increased efficiency on the network and shorter time-to-action. As well, the dashboard allows network domain experts to define engineering guidelines, manage the KPI definition, create predefined performance alerts, and define customized views for monitoring, troubleshooting, and issue research.

Developed for agricultural operations, Tupl Agro offers updated, objective, and personalized information for a farm operation and through an agricultural season, including alerts about farms, risk warnings, and agronomic recommendations. The information is delivered over WhatsApp to allow farmers to receive updates as they work. Further, the platform offers users historical data and reports, action planning, management of alerts and risks, and notification control to better manage their fields and operations.

Built for manufacturing, Tupl's AI Quality Control Toolkit is intended to help reduce the manual labor required to detect quality problems in manufacturing through the use of artificial intelligence and monitoring solutions. The platform is developed to allow users to create AI models for different parts of a manufacturing operation, while the platform uses Convolutional Neural Networks to detect and classify capabilities and quality problems. This allows the platform to actively learn and facilitate further automation throughout a manufacturing plant.

For managers, the platform offers individual and aggregated statistics at different levels—such as from production level statistics to company-wide statistics—to help users understand the capabilities of each manufacturing facility and production line while offering opportunities to increase efficiencies.