Maximize Your Productivity. Minimize Your Effort.
Harness the power of AtomAssist by NuclearN, where high-value tasks take precedence. Utilizing a chatbot interface, AtomAssist intuitively automates routine activities and provides insightful answers to user inquiries, streamlining operations with unparalleled precision and efficiency.
AtomAssist uses the latest in Generative AI techniques to be able to deliver nuclear related information with high accuracy and source references to facilitate compliance checks.
Streamlining routine tasks not only lowers internal costs but also boosts efficiency, allowing for more focused oversight on critical safety protocols.
Automate mundane tasks and improve documentation and compliance, enabling talent to innovate.
AI-driven platform enables fast uploading, searching, and referencing, optimizing research and collaboration.
Licensing Manager at a Nuclear Utility
“AtomAssist is helping us identify any plant related impacts to our licensing basis. This is an intenstive document search that would normally take a trained Engineer many hours - now it's performed in an instant. This is real AI accelerating our people.”
nuclear training texts
applicability
AtomAssist does not connect or interact with any public AI model or service provider. In other words, it does not use any API provided by a third party. We use a privately trained foundation model we created using large amounts of public Nuclear data. It is the world’s first private, nuclear-specific generative AI model. The current version is being developed and run at NuclearN corporate headquarters. Customers can choose whether to use NuclearN’s deployment, deploy on-premises, or host on their private cloud.
AtomAssist can be configured for both on-premises servers or cloud-based systems, in both CPU and GPU-accelerated environments. AtomAssist can be configured for low-workload use cases on a CPU-based system or scaled out on private GPU-accelerated hardware for high-workload use cases.
Our foundation LLM model (on which AtomAssist is based) is trained with many common plant designs currently in service as well as general physics and other technical topics, giving it a strong engineering context capable of answering many questions. Furthermore, we offer the ability to upload documents, which gives the Foundation LLM additional context. For up to several hundred documents, the model will retain high-quality results. For several thousands of documents, the model will need to be fine-tuned to gain context. If you wish to have thousands of documents loaded into the model for context, fine-tuning is required to optimize the embedding and retain high-quality results. This requires access to your data and 45 days to train and fine-tune a model specific to you.
Yes, the NuclearN platform exposes common REST APIs for all operations, which means it can integrate into a variety of EAM systems. We have performed integrations into Maximo, SAP, and Oracle EAM systems. We can handle the integration for you or work with an integration team on your side.
Yes, NuclearN has reporting functionality. For certain use cases, NuclearN can generate reports based on your data in MS Word or PDF formats. For all other use cases, NuclearN seamlessly connects to PowerBI for full reporting capabilities.
NuclearN can be deployed on-premise, on a customer's private cloud, or in Azure Government Cloud. There is no difference in platform functionality among the deployment environments. The Nuclear industry uses Azure Government Cloud for cloud deployments, which is FED-RAMP high and DOE 810 compliant and is where many Nuclear Utilities choose to host their cloud-deployed software applications.
We have 3 licensing options: 1. Active User (concurrent use) 2. Named Users 3. Enterprise (unlimited license) Prices are based on sizing. To get a feel for capability, we recommend starting with an annual subscription of 5 – 10 Active Users using our foundation model with RAG (document upload). We offer a seamless true-up model, so you can expand the license to fit your needs as usage and demand grows. We will provide monthly data on usage and a quarterly True-Up to actual usage (only pay for what you use).
For AtomAssist, the system requirements are as follows. For Azure: NV36 ads v5-series For AWS: EC2 G5.2 xlarge Instance For On-Premise: RedHat 8 or newer (RHEL), 32GB of RAM, 120GB free disk space, 8 Core intel CPU must support AVX instruction set, NVIDIA A10 or newer with minimum 24GB of VRAM.
There are 2 user interfaces (web browser and MS Word Add-in) that are agnostic of the operating system.
AtomAssist: Users have two main methods for data entry. Web Browser Interface: Users can input text directly into the chat interface or upload documents (including PDF, DOC, text files, and CSV) by dragging and dropping them into the platform. MS Word Add-in: Users can type text within an MS Word document. AtomAssist is capable of reading and interpreting the text directly from MS Word, providing a seamless integration. NuclearN Automation Products: For our automation solutions, we enable direct integration with your existing source systems such as Maximo and SAP, through REST APIs. This approach ensures efficient data exchange and process automation, enhancing system interoperability and operational efficiency.
Yes, NuclearN products offer seamless integration with PowerBI. By leveraging the NuclearN Platform's data APIs, you can easily export your data directly into PowerBI as a web data source. This integration allows for comprehensive data analysis, enabling unlimited reporting capabilities, trend analysis, charting, and more within PowerBI.
