DOE · DOE-530 · record R3387

Merlin - KCNSC Generative AI with RAG

Department of Energy · KCNSC - Kansas City National Security Campus (KCFO)

Overview

Development stage: PilotSource: b) Pilot – The use case has been deployed in a limited test or pilot capacity.
High-impact designation: Not high-impactSource: c) Not high-impact
Impact justification: Output does not serve as a principal basis for decisions or actions that have a legal, material, binding, or significant effect on rights or safety. Assists in developing software code that may accept product, but this AI will not make those decision
Start date: 2025-07-16 (day precision)Source: 2025-07-16T00:00:00
Withheld from public reporting: NoSource: a) No

Mission

Topic area: Information Technology
Operational functions: Search and knowledge retrieval Derived by keyword rules; see methodology

Problem the AI is intended to solve

Intended to address the opportunity to enhance overall productivity. No definite 'problem' being solved, just capitalizing on an opportunity to leverage industry investment in generative AI broadly.

Expected benefits

It serves as a general productivity enhancer -not driven by a specific problem, but by the opportunity to improve efficiency.

System outputs

outputs are context-aware responses that combine generated content with retrieved, authoritative information to ensure accuracy, relavance, and grounding.

Technology

AI classification: Generative AISource: Generative AI: AI that generates new or synthetic content (e.g., images, videos, audio, text, code).
System name(s): vLLM - OpenSource model engine OpenShift - containerization platform GPT-OSS-120b - Opensource model OpenWebUI - Opensource front end
Custom-developed code: No
Public source code: huggingface.co ↗

Sourcing

How it was built: Developed in-houseSource: b) Developed in-house
Vendor (as reported): N/A - Open Source Integration
Vendors (standardized): None on the standard list

A vendor is the supplier named by the agency. It does not identify the underlying model or AI technology.

Data and privacy

Involves PII: No
Privacy Impact Assessment: Not reported
Authorization to Operate: Yes
Demographic features: k) None of the above
Training and evaluation data: Leveraging OpenAI open-sourced models, so they are responsible for providing training data
Federal Data Catalog: Not reported

Governance

0 of 8 minimum-practice questions answered. A blank answer means the agency reported nothing; it does not mean the practice is absent.

Pre-deployment testingNo answer reported

AI impact assessmentNo answer reported

Independent reviewNo answer reported

Ongoing monitoringNo answer reported

Operator trainingNo answer reported

Fail-safeNo answer reported

Appeal processNo answer reported

User and public consultationNo answer reported

Potential impacts and how they were identified

Not reported