Skip to main content
AI & Platform Engineering

Building AI
systems that
ship.

Seven years of production AI — RAG pipelines, LLM infrastructure, data platforms, and cloud architecture for Fortune 500 firms and YC-backed startups. Available for consulting engagements via Innovate Loop.

7+
Years in production AI
300k+
Cases analyzed with AI
$5M+
Projected annual savings
3.5TB+
Data processed on platforms built
Explore

From prototype to
production infrastructure.

I work across the full stack of modern AI engineering — from designing retrieval pipelines to deploying hardened Kubernetes infrastructure. Engagements are project-based, fractional, or advisory.

01

RAG & LLM
Infrastructure

Design and build Retrieval Augmented Generation (RAG) systems that work in production — vector search, multi-signal reranking, and context optimization. Tooling via Model Context Protocol (MCP) where applicable. Built on real workloads, not demos.

Azure OpenAI Anthropic Claude Embeddings RAG MCP
02

Data Platform
Engineering

End-to-end data infrastructure for ML workloads — EHR pipelines, enterprise data lakes, and analytical platforms on MongoDB, PostgreSQL, and Snowflake. Built for scale, compliance, and real cost savings.

MongoDB Snowflake PostgreSQL ETL Python
03

Cloud & Kubernetes
Architecture

Production-grade cloud infrastructure with Kubernetes at the core. Multi-cloud deployments across Google Cloud, AWS, and Azure — from cluster provisioning to zero-downtime CI/CD. Kubestronaut-certified: CKA, CKAD, CKS, KCNA, and KCSA.

Kubernetes GCP AWS Azure Terraform Helm

Engineering at
enterprise scale.

I've shipped AI infrastructure for some of the most data-intensive organizations in finance and healthcare — and built core systems for YC-backed startups from day one. Each engagement is defined by real outcomes, not deliverable counts.

Charles River Development / State Street
AIFintech

AI pipeline across 300,000+ enterprise support cases

Designed and built embedding pipelines on Azure OpenAI to automatically classify support tickets, surface suggested next steps, and enhance documentation for future resolution — built against a dataset spanning 300,000+ total cases at ~40,000 monthly volume. Demonstrated measurable reduction in manual triage load and a clear path to production for State Street's global operations.

300k+
cases analyzed
Ardent AI — YC X26
RAGStartup

Core RAG infrastructure for a YC-backed AI startup

Built the foundational retrieval infrastructure from the ground up — hybrid vector search combining dense embeddings with BM25 sparse retrieval, multi-signal reranking, and context window optimization. Designed for rapid iteration without sacrificing production reliability. Delivered as part of the company's early technical foundation ahead of their YC batch.

YC
X26 batch
Blue Shield of California
DataHealth

3.5TB EHR data platform — $5M+ projected annual savings

Architected and built a large-scale electronic health records data platform on MongoDB and Snowflake, consolidating 3.5TB of fragmented patient data into a unified, query-optimized system. The platform enabled downstream ML workloads, improved reporting velocity, and was projected to generate $5M+ in annual operational savings through reduced manual data handling and improved clinical workflow efficiency.

$5M+
projected annual savings
Microsoft / Nuance
PlatformHealth

Kubernetes infrastructure for Dragon Medical Virtual Assistant — 400,000+ clinical users

Architected and scaled the Kubernetes infrastructure powering Dragon Medical Virtual Assistant, Nuance's AI-driven clinical documentation product used by over 400,000+ physicians and clinical staff globally. Focused on reliability, uptime, and zero-downtime deployment pipelines at clinical-grade scale.

400k+
clinical users

The tools
I work with.

Searchable by technology — if you need a specific stack, this is where to check. I work across the full AI engineering lifecycle, from data ingestion to model serving to production observability.

AI & LLM
LangChain Azure OpenAI OpenAI API Anthropic Claude RAG Vector Search MCP
Data Engineering
Python MongoDB Snowflake PostgreSQL Microsoft SQL Server MySQL Redis ETL Pipelines
Cloud & Infrastructure
Kubernetes Helm Azure Amazon Web Services Google Cloud Terraform Docker
Languages & Frameworks
Python TypeScript JavaScript Go FastAPI Node.js Express.js React / Next.js

Built on a deep
technical foundation.

Certification

Kubestronaut

Holder of all five CNCF Kubernetes certifications — CKA, CKAD, CKS, KCNA, and KCSA. The rarest tier of Kubernetes expertise.

Education

Cornell Tech, M.Eng. Computer Science

GPA 4.0 · Merit Scholarship · Startup Award Finalist. Graduate program at Cornell's NYC campus, focused on applied machine learning and systems engineering.

Certification

MongoDB Certified Developer & DBA

Dual-certified in MongoDB application development (C100DEV) and database administration (C100DBA) — relevant for data platform engagements at scale.

Research

Published — American Heart Association

Research published in Stroke, Vol. 55, No. 3 — demonstrating applied ML and NLP methodology in a clinical research context alongside domain experts.

Have a question?

Primarily production AI systems — RAG pipelines, LLM infrastructure, data platforms, and cloud architecture. I work best on engagements where real systems need to ship, not just prototypes. I've worked across fintech, healthcare, and early-stage startups. If the problem involves wiring together AI components into something that has to work reliably at scale, it's probably a good fit.

Both. I've worked with YC-backed startups building core infrastructure from day one, and Fortune 500 firms scaling existing systems. The work looks different — startups need speed and flexibility, enterprises need compliance and reliability — but the underlying engineering is the same. I adapt to the context.

Kubestronaut is a designation from the Cloud Native Computing Foundation (CNCF) awarded to engineers who hold all five Kubernetes certifications: CKA (administrator), CKAD (developer), CKS (security), KCNA (associate), and KCSA (security associate). It's the rarest tier of Kubernetes expertise — signaling deep, validated knowledge across the entire ecosystem, not just familiarity.

Yes. I work on project-based, fractional, and advisory engagements depending on what the work requires. If you need a senior AI engineer embedded on your team for a few days a week, or a focused project sprint, reach out and we can talk through what makes sense. All consulting is handled through Innovate Loop.

An AI platform engineer bridges the gap between ML research and production software. I build the infrastructure that lets AI systems run reliably — retrieval pipelines, model serving, data ingestion, observability, and the Kubernetes clusters that hold it all together. It's less about training models and more about making sure the entire system works at scale, handles failure gracefully, and can be iterated on quickly.

Let's build something
worth shipping.

Whether you need a RAG pipeline that actually works, data infrastructure built for scale, or a senior AI engineer embedded in your team — reach out.