AI Systems Engineer · Frederick, MD

Claire
Lindstrom

LLM Systems · Voice AI · RAG · Cloud Deployment

I build production AI systems.

About

Production AI systems engineering

I build and deploy production AI systems that operate reliably in real-world environments. My work spans LLM orchestration, agentic systems, RAG infrastructure, evaluation frameworks, and real-time voice AI platforms.

I specialize in the full stack of AI infrastructure: LLM systems and agents, semantic retrieval and RAG, evaluation and testing, and real-time communication systems.

My approach treats AI systems as distributed systems: modular orchestration layers, rigorous evaluation of LLM failure modes, and observability-first design.

Based in Frederick, MD · B.S. Computer Science (GPA: 3.49), UMGC 2025 · Certificate: Artificial Intelligence Foundations · U.S. Citizen, clearance eligible

Projects

Production AI systems I've built

LIVE · FIELD OPS
TwilioOpenAI RTFastAPIAWS
Employer Project

Voice AI for Field Ops

Production System

Real-time voice orchestration system for field operations. Built production-grade infrastructure with reliability and latency optimizations, AWS deployment, and integrations with Twilio and OpenAI Realtime API.

What I Owned

  • Architected real-time audio streaming pipeline
  • Implemented session lifecycle and state management
  • Optimized latency and reliability metrics
  • Built monitoring and observability systems
  • Deployed and maintained AWS infrastructure
TwilioOpenAI RealtimeFastAPIWebSocketAWS EC2Redis
Manifest Alchemy AI
Founder Project

Manifest Alchemy AI

Founder Project

Agentic planning platform with personalized AI memory, structured reasoning, and RAG-powered retrieval. Built with Next.js, Supabase, and vector search for intelligent task orchestration.

Next.jsReactPostgreSQLSupabaseVector DBLLM Orchestration
Document Processing AIFastAPI
report.pdf
notes.txt
data.pdf
Query your documents...FAISS
Personal Project

Document Processing AI

Personal Build

Production-grade RAG API built with FastAPI. Upload PDFs and plain-text files, then query them in natural language. Features FAISS vector indexing, character-based chunking with word-boundary snapping, SQLite metadata persistence, and an abstracted LLM provider pattern.

PythonFastAPIFAISSSQLiteRAGLLM Abstraction
Generated Moon AI Design Agency
Founder Project

Generated Moon

Luxury AI Design Agency

Full-stack AI-powered design agency platform. Features an intelligent chatbot that conducts a structured assessment (5–8 questions) to auto-generate detailed project briefs, Stripe payments, Resend email integration, and lead capture — all deployable on Railway or Vercel.

ReactViteExpressClaude AIHuggingFaceStripeResend
Synergy Living
Founder Project

Synergy Living

Wellness Platform

Multi-page wellness platform built with clean semantic HTML, custom CSS, and vanilla JavaScript. Covers Home, About, Resources, Newsletter, and Stories — a full brand identity and content system with zero dependencies.

HTMLCSSJavaScriptResponsive DesignUI/UX

Skills

Technologies and capabilities

Languages

  • Python
  • JavaScript
  • SQL
  • Java
  • C++

AI & ML

  • OpenAI APIs (Realtime & Chat)
  • LLMs
  • NLP
  • Prompt Engineering
  • AI Evaluation
  • TensorFlow

Web & Backend

  • FastAPI
  • React
  • Next.js
  • HTML
  • CSS
  • Tailwind
  • WebSockets

Databases

  • PostgreSQL
  • Supabase
  • SQLite
  • Airtable
  • Pinecone
  • Vector Databases

Cloud & DevOps

  • AWS EC2
  • Load Balancing
  • Git/GitHub
  • Docker

Experience

Professional background

AI Software Engineer (Voice AI Systems)

Previous Employer
2025 - Present

Architected and deployed production-grade Voice AI system using Twilio Media Streams, OpenAI Realtime, FastAPI, and AWS EC2. Built low-latency WebSocket audio pipelines for real-time transcription and response generation. Implemented role-based conversational logic and dynamic data retrieval from operational databases. Designed caching and pre-generation strategies to eliminate cold-start latency and call drop-offs. Debugged and resolved real-time call reliability issues in a live production environment.

Founder & AI Engineer

Manifest Alchemy
2024 - Present

Designed and built full-stack AI platform focused on long-term memory and personalized reasoning. Implemented LLM-powered conversational agents with structured prompts and semantic retrieval. Built modern web interfaces using React, Next.js, and Tailwind CSS. Designed scalable data models using PostgreSQL/Supabase with vector storage.

AI & Front-End Software Engineer

Re:Members
2024 - 2025

Developed and maintained JavaScript, HTML, and CSS front-end components. Wrote and optimized SQL queries supporting customer and operational workflows. Assisted in training and evaluating AI chatbot systems, improving accuracy and reliability. Performed AI quality assurance, identifying hallucinations and failure cases prior to deployment.

Data & AI Solutions Specialist

The Kelly Companies
2022 - 2023

Developed AI-driven VR/AR training programs using Unity, Unreal Engine, and Python. Built data-rich web platforms using React, Python, and SQL for analytics and LMS integration. Created AI-powered learning materials using ChatGPT API and Pinecone vector databases for semantic retrieval. Automated production workflows with Python, increasing efficiency by 40%. Contributed to AI-assisted visualization tools for training and performance tracking.

Prototypes

Interactive demos and experiments

SmartSiteVOICE AI
Your website,
built by voice
"Add testimonials""Dark theme"
React · Vite · Web Speech API
Interactive Prototype

Smart Site

Voice AI Web Builder

Speak commands to modify a live website in real time. "Add testimonials", "dark theme", "add booking section" — the AI interprets voice input and updates the layout instantly with animated transitions.

ReactViteWeb Speech APIClaude APIOpenAI APICSS
"Add testimonials""Dark theme""Add booking section""Make it modern"
View Details