Post Job Free
Sign in

AI Engineer: Production RAG & LLM Platforms

Location:
Houston, TX
Posted:
November 11, 2026

Contact this candidate

Resume:

JAMES BORLAND

AI ENGINEER RAG, LLM APPLICATIONS & CLOUD PLATFORMS

Hoover, AL 432-***-**** **************@*****.*** linkedin.com/in/james-borland-689950430 PROFESSIONAL SUMMARY

Hands-on AI Engineer with 12+ years of software engineering experience building backend, cloud, data, and AI applications. Experienced in production RAG pipelines, LLM-powered workflows, vector search, asynchronous APIs, and multi-tenant SaaS using Python, FastAPI, LangChain, LlamaIndex, PostgreSQL, Pinecone, AWS, and Terraform. Built AI-assisted document workflows that reduced manual compliance review by 40%. Strong focus on production coding, retrieval quality, LLM cost optimization, debugging, observability, and reliable deployment. TECHNICAL SKILLS

AI / LLM: RAG, LangChain, LlamaIndex, OpenAI APIs, Pinecone, pgvector, embeddings, vector search, prompt engineering, metadata filtering, retrieval tuning

Backend: Python, FastAPI, Node.js, Spring Boot, Django REST Framework, REST APIs, async/await, background processing

Data: PostgreSQL, SQL, data modeling, ETL, data validation, document ingestion, chunking, vector indexing Cloud / DevOps: AWS ECS, Lambda, API Gateway, Terraform, Docker, CI/CD, Infrastructure-as-Code Observability: Prometheus, Grafana, application logging, metrics, production troubleshooting Frontend / Architecture: React, TypeScript, JavaScript, multi-tenant SaaS, RBAC, distributed applications PROFESSIONAL EXPERIENCE

Senior AI Platform Engineer NuSkin Dec 2022 – Present Provo, UT

• Built and deployed production RAG pipelines using Python, FastAPI, LangChain, LlamaIndex, Pinecone, and AWS, processing approximately 25,000 supply-chain documents per month and reducing manual compliance review by 40%.

• Developed end-to-end document workflows covering ingestion, parsing, chunking, embedding generation, metadata extraction, vector indexing, retrieval, prompt construction, and LLM response generation, supporting approximately 12 document types and approximately 150 internal users.

• Implemented Python async/await and FastAPI background processing for long-running ingestion and LLM workloads, increasing concurrent job throughput by approximately 40% and keeping API response times below 2 seconds for non-inference requests.

• Improved retrieval precision by approximately 25% through chunk-size experiments, metadata filters, query constraints, and LlamaIndex retrieval tuning, reducing irrelevant context sent to the LLM.

• Built RAG evaluation and debugging workflows to inspect retrieved chunks, source relevance, prompt inputs, model responses, and failure logs, reducing investigation time for inaccurate responses by approximately 35%.

• Applied metadata pre-filtering before vector retrieval, reducing average prompt-token usage by approximately 30% and lowering monthly LLM inference cost by approximately 20%.

• Containerized 8+ Python and FastAPI services with Docker and provisioned repeatable AWS environments with Terraform, reducing environment setup time from 2 days to 3 hours.

• Implemented Prometheus and Grafana monitoring across 10+ services, tracking API latency, background-job failures, retrieval errors, and infrastructure health; reduced mean time to detect issues by approximately 40%.

• Diagnosed and resolved 10+ monthly production issues across REST APIs, vector retrieval, PostgreSQL, asynchronous processing, AWS infrastructure, and third-party LLM integrations, maintaining 99.5% service availability.

JAMES BORLAND

Hoover, AL 432-***-**** **************@*****.*** linkedin.com/in/james-borland-689950430 PROFESSIONAL EXPERIENCE CONTINUED

Full Stack Software Engineer HelloTax Nov 2020 – Dec 2022 Alicante, Spain

• Built and maintained Node.js and Spring Boot services for a multi-tenant European VAT compliance platform supporting 1,000+ customers and processing 5 million+ transaction records annually.

• Implemented tenant isolation with discriminator columns, repository-level query scoping, JWT authentication, and RBAC across 20+ APIs, with zero confirmed cross-tenant data incidents.

• Developed Spring Boot services for financial VAT calculations and Node.js services for high-concurrency I/O, asynchronous workloads, and 8+ third-party integrations.

• During later project phases, contributed to an AI-assisted legislative-summary prototype by preparing structured source context and citation requirements.

• Improved PostgreSQL performance during peak transaction processing by designing composite indexes and asynchronous workflows, reducing selected query latency by approximately 45% and batch-processing time by approximately 30%.

• Built API validation, error handling, and integration logic for high-volume financial workflows, reducing failed or invalid transactions by approximately 20%.

• Resolved 15+ production defects per quarter across APIs, SQL queries, authentication, tenant scoping, asynchronous jobs, and third-party integrations.

Full Stack Software Engineer Altira Health Feb 2017 – Oct 2020 Birmingham, AL

• Developed healthcare analytics applications using Express.js, React, REST APIs, and WebSockets to process 50,000+ patient-feedback and clinical-survey records.

• Integrated NLP-based sentiment analysis and theme identification into backend workflows, reducing manual feedback categorization effort by approximately 35%.

• Built 25+ backend API endpoints, validation rules, data integrations, and React components supporting healthcare reporting and analytics workflows.

• Implemented JWT authentication and role-based access controls for services handling sensitive healthcare information, supporting HIPAA compliance requirements.

• Used WebSockets to deliver dashboard updates within 2 seconds, reducing reliance on manual refresh workflows for 200+ users.

• Troubleshot issues across frontend components, REST APIs, authentication, data integrations, and NLP processing pipelines.

Full Stack Software Engineer Leidos May 2014 – Jan 2017 Reston, VA

• Developed enterprise healthcare reporting applications using Django REST Framework, Python, Node.js, and REST APIs for 500+ internal users across multiple business units.

• Built data integration and normalization pipelines processing 100,000+ records per batch from 6+ enterprise systems, validating inconsistent patient identifiers and clinical metrics.

• Implemented validation and error-handling logic that reduced malformed records entering centralized reporting workflows by approximately 25%.

• Containerized applications with Docker to improve environment consistency and managed runtime configuration to keep sensitive settings outside source code.

• Developed 40+ Jest integration tests for API and data-processing integration points, increasing critical-path test coverage to approximately 80% and reducing ETL regressions by approximately 30%. EDUCATION

Bachelor’s Degree, Computer Science The University of Alabama Sep 2011 – May 2014



Contact this candidate