Vetted Latency Optimization Professionals

Chase Bradley

Screened ReferencesModerate rec.

Staff AI Engineer specializing in LLM systems, retrieval, and ML infrastructure

Remote10y exp

“ML/LLM engineer from Cohere who has owned retrieval, reranking, agentic workflows, and internal evaluation infrastructure end-to-end in production. Particularly strong in turning brittle RAG and research-heavy ideas into scalable enterprise systems with grounded outputs, lower customer escalations, and adoption by major clients like Notion and Fujitsu.”

“Senior engineering/technology leader with experience driving board-backed, multi-year platform transformations in retail commerce (American Eagle) and scaling ML/product delivery platforms at Amazon Alexa. Known for data-driven build-vs-buy decisions, org scaling (6→80), and measurable outcomes including 98.5%→99.9% uptime, 234 experiments generating ~$200M revenue lift, and increasing delivery velocity from 1.5 to 30 features/week while transitioning to DNN/large-model architectures.”

Christopher Sipola

Screened

Senior Machine Learning Engineer specializing in LLMs and recommendation systems

New York, NY11y exp

“ML/GenAI engineer who owned major parts of Spotify’s AI DJ from offline experimentation through deployment, monitoring, and iteration. They combine recommender systems, RAG, real-time feedback loops, and LLM safety/orchestration to ship consumer-facing personalization features that drove double-digit engagement and deeper listening sessions.”

“Built and deployed a production LLM conversational AI system at OpenAI supporting chat, summarization, and semantic search at 1M+ requests/day, driving major latency (40%) and accuracy (25%) improvements through Pinecone optimization and tighter RAG with re-ranking. Also has Amazon experience improving recommendation systems by translating ML metrics into business terms to boost CTR and conversions, with strong MLOps/orchestration depth (Airflow, MLflow, SageMaker, Kubeflow).”

“ML engineer with 5 years at Stripe building and productionizing real-time fraud detection at massive scale (3M+ transactions/day; $5B+ annual payment volume). Delivered measurable impact (22% accuracy lift, 18% loss reduction, +3–5% authorization rates) and has strong MLOps/orchestration experience (Docker, Kubernetes, Airflow, MLflow, CI/CD, monitoring/rollback) plus a structured approach to LLM agent/RAG evaluation.”

“Co-founded Aethernet Inc., a B2B IoT connectivity platform with a usage-based revenue model and direct sales motion to device manufacturers. Bootstrapped with founder capital and strict cost control, with angel interest pending technical validation; developed the product through multi-year architecture experiments to deliver measurable latency/reliability/scalability/cost advantages versus DevOps-heavy incumbents like AWS IoT Core.”

“LLM/agentic systems practitioner who partners directly with customers to productionize prototypes end-to-end—defining business-aligned metrics, building evaluation datasets, and shipping monitored, cost-bounded inference APIs on AWS Lambda. Notably delivered a vehicle damage classification system that cut manual review by 40% and stabilized agent workflows by instrumenting state transitions to uncover and fix a race-condition-driven skipped tool call.”

Nicholas Ogura

Screened ReferencesStrong rec.

Executive Platform & Security Engineering leader specializing in multi-cloud Kubernetes and FinTech

San Diego County, USA18y exp

“Startup-focused infrastructure/security leader who stepped into head of engineering and re-platformed an entire product end-to-end in 3 months to meet launch. In crypto/fintech, recognized the market-data system as an ETL/data product and rebuilt it as a separable, securely accessible platform—prompting inbound interest within a week—while advocating an open-source-first observability stack (Prometheus/Grafana/Loki) to avoid vendor lock-in.”

“NVIDIA engineer who built and shipped a production LLM-powered enterprise knowledge system (summarization, transcription, and Q&A) that cut document retrieval time ~30%. Deep hands-on experience with RAG (FAISS/Pinecone), GPU-accelerated microservices on AWS, and reliability/safety practices (Guardrails AI, prompt A/B testing, canary releases) plus strong MLOps orchestration across Airflow, Step Functions, and Kubernetes GitOps.”