Technical Architect - ML - GenAI

Quantiphi

21d•Remote

About The Position

Quantiphi is seeking a hands-on Generative AI Architect / Lead to design and deliver enterprise-grade GenAI solutions using AWS Bedrock and Agentcore. The role focuses on scalable applications involving large language models (LLMs), retrieval-augmented generation (RAG), and agentic AI workflows. The architect will define solution architecture, guide teams, contribute to development, and ensure performance, scalability, and cost efficiency. The position is remote in the US and calls for 8+ years of experience.

Requirements

At least 8 years of relevant hands-on technical experience implementing and developing cloud ML solutions on AWS.
Hands-on experience with AWS services, including SageMaker and Bedrock, across different data sources, training jobs, and real-time and batch applications.
Experience designing and implementing agentic AI architectures using frameworks such as LangChain or Strand Agents.
Hands-on experience with Amazon AgentCore for building, deploying, and scaling production-grade agentic AI applications, including agent memory management, tool registry, and observability.
Experience architecting and deploying scalable AI solutions on AWS, including Lambda, Bedrock, Step Functions, S3, API Gateway, and SageMaker.
Proficiency with LLM APIs, such as Claude, Nova, and other third-party LLM providers, including API integration and multi-model orchestration strategies.
Hands-on experience fine-tuning or optimizing large language models.
Familiarity with LLM tool use, prompt templating, and context management.
Strong expertise in vector databases, including indexing strategies, embedding generation, similarity search, and integration with RAG architectures.
Experience evaluating LLM zero-shot and few-shot capabilities, tuning fine-tuning hyperparameters, assessing task generalization, and exploring model interpretability for web application integration.
Experience developing and maintaining MCP implementations for state, context windows, memory, and prompt orchestration across distributed agent systems.
Experience with at least one workflow orchestration tool, such as Airflow, Step Functions, SageMaker Pipelines, or Kubeflow.
Experience implementing secure, scalable APIs and integrating third-party data sources and tools.
Ability to collaborate with cross-functional teams and stakeholders to understand requirements and implement solutions.
Experience with deep learning concepts, including transformers, BERT, attention models, tokenization, and embeddings.

Nice To Haves

Experience with software development and exposure to frontend and backend frameworks and communication protocols.
Experience with Infrastructure as Code (IaC) and CI/CD pipelines.
Experience with NLP concepts such as syntactic and semantic analysis and named entity recognition (NER).

Responsibilities

Design and implement GenAI solutions using AWS Bedrock and Agentcore.
Define architecture for LLM-based applications, including RAG pipelines and agentic workflows.
Develop and orchestrate agentic AI workflows for multi-step reasoning, tool use, and task automation.
Build and manage RAG pipelines, including embeddings, retrieval mechanisms, and vector databases.
Integrate LLM capabilities into enterprise applications through APIs and backend services.
Design and optimize prompt engineering strategies for accuracy, relevance, and performance.
Work with structured and unstructured data sources to enable knowledge-driven AI applications.
Evaluate, monitor, and optimize models for latency, cost, and response quality.
Collaborate with application, data, and platform teams to deliver end-to-end solutions.
Define best practices for security, governance, and responsible AI use.
Troubleshoot and resolve issues in production GenAI systems.
Provide technical leadership and mentor team members while remaining hands-on.
Design and deploy scalable AWS AI solutions using services such as Lambda, Bedrock, Step Functions, S3, API Gateway, and SageMaker.
Develop and maintain Model Context Protocol (MCP) implementations to manage state, context windows, memory, and prompt orchestration across distributed agent systems.
Implement secure, scalable APIs and integrate with third-party data sources and tools.
Collaborate with developers, QA, project managers, and other stakeholders to understand requirements and implement solutions.