01
Your team answers the same questions repeatedly.
I turn scattered tickets and documentation into self-service tools that return useful, source-backed answers.
CVS Health: 16% fewer support tickets and 20% faster resolution
Experience
Four years across CVS Health, Aetna, and Evidenza—building internal tools, customer-facing platforms, AI-assisted workflows, and the safeguards needed to operate them.
16%
fewer support tickets
Self-service answers from trusted company sources
20%
faster issue resolution
Less time spent searching past tickets and documentation
75%
less manual upkeep
Workflow automation for engineering teams
14%
higher engagement
Better signals for ad recommendations
Where I help
I am most useful when the problem crosses product, engineering, and applied AI—and someone needs to own the path from ambiguity to a working system.
01
I turn scattered tickets and documentation into self-service tools that return useful, source-backed answers.
CVS Health: 16% fewer support tickets and 20% faster resolution
02
I map the handoffs, automate the repetitive steps, and build a clear interface around the work people still need to own.
CVS Health: 75% less manual upkeep across four teams
03
I add monitoring, useful operational views, and safer delivery practices so teams can see failures and respond sooner.
CVS Health: 12% less downtime and about 25% faster recovery
04
I build evaluation into the workflow so generated content is measured, compared, and rejected when it is not an improvement.
CVS Health: 18% more catalog images with useful captions
Selected experience
The technical details are available, but the outcomes come first.
Graduate AI internship · Jan–May 2026
Brooklyn, NY · During M.S. studies at NYU
The problem
Customer profiles missed less-common traits, and creative teams needed a repeatable way to assess competitor ads.
What I built
A profile generator that improved its own instructions through scored feedback, plus an ad-analysis workflow using both images and text.
What changed
80% coverage of long-tail traits
14% engagement lift in A/B testing
2023–2025 · Promoted in 2024
CVS Health · New York, NY
Problem: Support knowledge was trapped in tickets and internal documentation.
Built: Built a self-service assistant that answered recurring questions from those trusted sources.
16% fewer support tickets and 20% faster resolution.
Problem: Product images were missing useful, on-brand captions.
Built: Built an AI-assisted caption workflow with an automatic quality check. New text was saved only when it improved on the existing caption.
18% more catalog images had useful captions.
Problem: Engineering workflows required repeated status, assignment, and board updates.
Built: Built the planning view and automation used across four teams.
75% less manual upkeep.
Problem: Customer-facing failures were difficult to diagnose quickly.
Built: Owned on-call reliability and built monitoring that made incidents easier to see and investigate.
12% less downtime and roughly 25% faster recovery.
Feb 2022–Jan 2023
New York, NY
The problem
Engineering operations relied on repetitive quality checks, platform maintenance, and migration steps.
What I built
Internal automation, data-migration workflows, validation tools, and platform health checks using Node.js and MongoDB.
What changed
Less repetitive QA and safer, more consistent platform changes.
React, TypeScript, Node.js, Python, FastAPI, MongoDB, PostgreSQL, Google Cloud, Terraform, GitHub Actions, monitoring with Grafana and OpenTelemetry, retrieval-based assistants, vision-language models, and automated AI evaluation.
Education
Courant Institute
M.S. Computer Science (AI)
B.S. Software Engineering & Mathematics
Have a problem like these?