Senior Generative AI Engineer

Quevera LLC is supporting an enterprise Generative AI optimization engagement for a large-scale customer in

Herndon, VA and Springfield, VA

(onsite). In this role, you will help modernize GenAI on

AWS , working embedded with customer Scrum teams to design and implement production-ready capabilities such as model routing and orchestration, conversation-context solutions, and integration-ready GenAI services. What you'll build

Design and implement

intelligent model-routing logic

using ML-based classification to route requests to appropriate LLMs based on conversational context, task complexity, and performance requirements. Build and maintain

orchestration layers

using

AWS Bedrock ,

LangChain , or custom frameworks to support

multi-model workflows . Train, evaluate, and iterate on

classification models

using evaluation frameworks and feedback loops. Architect

conversation-context analysis , retrieval, and embedding strategies to interpret user intent across

multi-turn interactions . Develop

prompt-engineering standards

and context-window management techniques that support intent classification and routing. Architect enterprise-scale GenAI solutions with

high availability ,

performance optimization , and

cost management . Apply GenAI platform practices including

observability ,

model evaluation ,

guardrails ,

responsible AI controls , and

versioning . Design and maintain

data extraction and transformation pipelines

supporting model training, evaluation, and enterprise application integration. How you'll work with the customer team

Engage with customer stakeholders and technical counterparts to gather requirements, validate solutions, and participate in

Agile/Scrum ceremonies . Collaborate with the Engagement Manager to track progress and risks, producing technical documentation, architecture diagrams, and handoff artifacts. Requirements

Active TS/SCI clearance with Polygraph required. 5+ years

of software development and cloud architecture experience. 2+ years

of hands-on

AWS

experience with production workloads. 2+ years

of experience with

Generative AI technologies

and

LLM deployment . Proven track record of

enterprise-scale

architecture design and implementation. Experience with

NatSec customers . Deep understanding of

GenAI foundation models

and deployment patterns. Strong programming skills in

Python

and infrastructure-as-code tools. Experience with

vector databases

and

semantic search technologies . Expert-level knowledge of AWS services, including

Amazon Bedrock ,

OpenSearch Service ,

CloudWatch ,

ECS/EKS , and

RDS/DynamoDB . Strong understanding of security automation and compliance frameworks. Ability to work independently and drive initiatives with minimal oversight. Strong problem-solving and analytical thinking capabilities, including leading technical discussions and influencing architecture decisions. Ability to work in ambiguous, fast-paced environments with a focus on delivering measurable customer outcomes. Must be able to obtain and maintain

Springfield, VA customer accounts . Must be able to travel onsite periodically for meetings and ceremonies. Technology focus

AWS ,

Amazon Bedrock ,

LangChain ,

Python ,

OpenSearch Service ,

CloudWatch ,

ECS ,

EKS ,

RDS ,

DynamoDB ,

vector databases ,

semantic search technologies Benefits

100% employer-paid medical coverage

(optional plan) Competitive options for

Medical, Dental and Vision

insurance Employer-paid short-term and long-term disability

coverage Employer-paid life insurance $5,000 annually

for education, training, certifications, and professional development Career advancement through the structured

iQTouch Program Up to

6% 401(k) match Additional

4% profit-sharing

contribution (company discretion) Desired skills

Ability to contribute to complex technical challenges and prototype innovative solutions. Ability to effectively transfer knowledge to customer teams. Ability to balance strategic thinking with hands-on implementation. Ability to deliver enterprise-grade solutions under tight timelines.

#J-18808-Ljbffr