AI Developer

AI Developer Remote (U.S.) • Full-Time Permanent About the Role My client is looking for a skilled AI Developer to build and advance AI/ML-intensive features for our customer platform. This is a hands-on engineering role focused on financial entity extraction, continuous learning pipelines, scenario modeling, and model evaluation — all powered by Amazon Bedrock. You will work at the intersection of prompt engineering, machine learning, and backend development, designing systems that continuously improve through human-in-the-loop feedback and rigorous evaluation frameworks. If you have deep experience building production AI pipelines on AWS and a strong command of Python, this role is built for you. What You'll Do Design and build AI/ML-intensive features for our customer platform including financial entity extraction, decision and action entity identification, and scenario modeling Engineer and iterate on prompts for Amazon Bedrock (Claude Sonnet/Haiku) to improve extraction quality, structured outputs, and Converse API interactions Develop and maintain continuous learning pipelines that incorporate human corrections to improve model performance over time Implement and manage RAG (Retrieval-Augmented Generation) patterns using OpenSearch Serverless vector stores and Bedrock Knowledge Bases Build and orchestrate multi-step AI workflows using AWS Step Functions Write Python 3.13 Lambda functions for data processing, model invocation, and pipeline automation Design and execute model evaluation frameworks including precision/recall measurement, extraction accuracy benchmarking, and golden test set development Build and maintain HITL (Human-in-the-Loop) feedback loops that use human corrections to continuously refine prompt performance Work within customer-managed Bedrock proxy constraints, ensuring all production AI interactions comply with governed access patterns Collaborate cross-functionally with product, data, and platform engineering teams to ship reliable, scalable AI features Required Skills & Experience Amazon Bedrock — hands-on production experience with Claude Sonnet and Haiku models, Converse API, and structured output generation Prompt engineering — deep expertise in designing, testing, and iterating prompts specifically for entity extraction in financial, decision, and action contexts Python programming — strong proficiency in writing Lambda functions, data processing pipelines, and AI workflow automation RAG architecture — practical experience implementing retrieval-augmented generation using OpenSearch Serverless vector stores and Bedrock Knowledge Bases Model evaluation — proven ability to design evaluation frameworks using precision/recall metrics, extraction accuracy measurement, and golden test sets AWS Step Functions — experience orchestrating complex, multi-step AI workflows in production environments HITL feedback loops — experience designing systems that capture human corrections and feed them back into prompt or model improvement cycles Managed Bedrock proxy environments — ability to operate within customer-governed Bedrock access constraints without direct production Bedrock access Nice to Have Experience in financial services or fintech platforms with exposure to financial document processing or data extraction Familiarity with Bedrock Agents and multi-agent orchestration patterns Experience with AWS CDK or Terraform for infrastructure-as-code in AI/ML environments Familiarity with LLM observability and monitoring tools Background in MLOps or continuous model improvement workflows Show more Show less