Application Architect

USDC•Plano, TX

3h•Onsite

About The Position

Genesis10 is seeking an Application Architect for a contract position with a Global Financial Institution. This role is responsible for training and maintaining machine learning models for a multi-channel Virtual Assistant. The ideal candidate will leverage Natural Language Understanding (NLU), Large Language Models (LLMs), and Agentic AI capabilities to deliver intelligent, context-aware conversational experiences across web, mobile, and voice channels.

Requirements

2+ years of experience with conversational interfaces, natural language processing, and conversational AI systems
Strong understanding of intent recognition, intent classification, entity extraction, semantic search, and dialogue management concepts
Experience training machine learning algorithms for data classification, speech recognition, and natural language understanding
Experience developing and evaluating text summarization solutions using Large Language Models (LLMs) and Generative AI technologies
Knowledge of Agentic AI frameworks, multi-agent systems, autonomous task execution, tool orchestration, and reasoning-based AI architectures
Experience with Python and AI/ML development libraries and frameworks
Unique skillset in computational linguistics combined with strong technical and analytical expertise
Working knowledge of LLMs, Generative AI, and conversational AI platforms
Familiarity with version control and development lifecycle tools such as Git, SVN, JIRA, and Azure DevOps
Experience working in DevOps and Agile environments
Strong analytical, troubleshooting, and problem-solving skills
Strong verbal and written communication skills

Responsibilities

Understand the intent portfolio for NLU across domains and how it maps to conversation design for web, mobile, and voice channels
Design, develop, and optimize intent recognition frameworks to improve user request classification, intent resolution, and conversational accuracy
Identify and build appropriate datasets to train and test machine learning models for intent classification, entity extraction, speech recognition, and semantic understanding
Develop tools and telemetry that can measure and monitor accuracy, performance, intent recognition quality, and model effectiveness
Develop and implement summarization capabilities using LLMs to generate concise, accurate summaries of user interactions and conversational outcomes
Develop a conversational AI strategy leveraging NLU, LLMs, and Agentic AI architectures to enable autonomous task orchestration and workflow execution
Work with agentic workflows that enable AI agents to reason, plan, retrieve information, invoke tools, and execute tasks
Develop disambiguation, clarification, and error-handling strategies as the virtual assistant scales
Monitor conversations and interaction analytics to identify underperforming intents, summarization gaps, and agent behaviors, and develop solutions to improve performance
Evaluate and optimize prompt engineering, retrieval-augmented generation (RAG), summarization quality, intent detection accuracy, and agent decision-making performance
Collaborate with data scientists, product owners, UX researchers, and engineers to continuously improve the virtual assistant