Open to full-time roles
I catch what does not add up, then check that the claim holds up, through dashboards, agents, and computer-vision tools.
Looking for roles as a Data Analyst, Analytics Engineer, or Data and AI Engineer. Currently Business Intelligence Analyst, Climate and ESG, at Worldview Development USA. NYC metro, remote.
I'm a data and AI analyst with four-plus years turning large, messy datasets into decisions across climate, healthcare, and public service.
I design BI dashboards and KPI frameworks in Power BI, Tableau, and Looker, automate ETL and ELT pipelines in Python, SQL, Snowflake, and dbt, and build machine learning and generative AI systems, from GenAI-powered NLP for text clustering to computer vision for image classification. Lately I've been just as interested in what happens after the dashboard ships: whether an agent's recommendation actually holds up, and whether a sustainability claim actually has a number behind it.
Every number here is a result I can walk you through.
21.5M
tonnes CO2e tracked across restoration programs
140K+
community members reflected in program reporting
$2M+
in grant budgets automated for variance reporting
480
hours of manual work eliminated every year ($38K saved)
40%
faster analytical reporting with a GenAI summarization workflow
30%
higher model recall on a 70M+ record predictive model
Mangroves are being lost at roughly a hectare every 15 minutes. The market meant to fund their protection has not kept pace. Here is what the registry data shows.
Each tile is one hectare. In real time, one disappears every 15 minutes. Here it happens every 1.2 seconds, 750 times faster.
0hectares lost in this demo
Of the 91 mangrove and coastal projects ever listed on the world's largest carbon registry, only 8 have ever issued a credit.
Filter by what you are curious about.
Showing all 6 projects
Climate
Merged NASA fire data, World Bank deforestation trends, and every credit ever issued on the Verra registry. Found that 83 of 91 projects have never issued a credit, and one project alone accounts for 81% of global supply.
Climate
A public dashboard tracking global mangrove forest loss, rebuilt on free, no-card infrastructure so it keeps running without a budget behind it.
Fashion AI
A computer-vision app that reads a photo of your closet and builds outfits from what you already own, so dressing sustainably beats buying new.
Climate AI
An AI agent that reads sustainability reports and scores every emissions claim against a 7-point credibility rubric, flagging the ones too vague to verify, with the exact sentence behind every flag. Tested on a real 83-page report, where it caught a Scope 3 increase missing from the highlights page.
Built for the AI Builders Hackathon, 2026
AI
An autonomous AI agent for data-incident response. Twelve incident scenarios, built and tested end to end, catching pipeline failures and unreliable agent behavior before they become outages.
Climate
Pulled 304,611 satellite methane observations over the continental U.S. from ESA's Sentinel 5P, compared them with NOAA background levels and official UNFCCC inventories, and found satellites are better at locating emissions than at verifying national totals.
Independent open-source research
These are the three things I check first in any sustainability claim: a real number, a baseline, and a defined scope. Pick a sample or paste your own sentence.
This is a simplified rule-based illustration of three checks. The real ScopeCheck uses the Claude API to extract claims from full reports and score each one against 7 criteria (baseline year, scope, absolute vs intensity, third-party verification, time-bound target, offset disclosure, methodology), citing the exact sentence behind every flag.
Results show up here.
Worldview Development USA
Design executive dashboards and KPI frameworks across climate, restoration, and grant programs, and automated $2M+ in grant budget variance reporting for C-suite and board-level decisions.
211 LA, County of Los Angeles
Built automated SQL-to-Redshift pipelines and GenAI-powered text clustering for disaster response reporting, eliminating 480 hours of manual work a year and cutting reporting time 40%.
Tri Towers, Kent State University
Built classification and regression models on 70M+ behavioral records, improving model recall 30% through feature engineering, hyperparameter tuning, and cross-validation.
Tesla Diagnostics Center
Built real-time anomaly detection across 1.8TB of patient vital-sign data, preventing 470+ critical device escalations across 14 clinics under HIPAA-compliant governance.
Fast fashion and unclear climate claims share the same failure: nobody checks the receipts.
I spend my days building the receipts. Closet First Stylist uses computer vision so people reach for clothes they already own instead of buying new. ScopeCheck uses an AI agent to catch sustainability claims that don't cite a real number. Mangrove Monitor and the methane tracker turn satellite data into something a policy analyst can actually read. Different fields, same habit: follow the data until the story holds up, or admit that it doesn't.
I work across finance and operations, climate, healthcare, AI and technology, and fashion and retail, and the same two instincts keep recurring: catch what does not add up, then check that the claim holds up.
Finance and operations
I automated $2M+ in grant budget variance reporting and built the executive dashboards and KPI frameworks that C-suite and board decisions rely on.
Climate
I showed that 83 of 91 mangrove and coastal projects on the Verra registry have never issued a credit, and that one project accounts for 81% of global supply.
Healthcare
I built real-time anomaly detection on 1.8TB of patient vital-sign data, preventing 470+ critical device escalations across 14 clinics under HIPAA-compliant governance.
AI and technology
I built an agent that scores emissions claims against a 7-point credibility rubric and, on an 83-page report, caught a Scope 3 increase missing from the highlights page.
Fashion and retail
I built a computer-vision app that reads a photo of your closet and builds outfits from clothes you already own.
I helped the Worldview Development USA team put together "Reimagining Mangroves and the Blue Economy in Panama" at Bryant Park Loft. If you are building at the intersection of climate, data, or fashion, I would still love to connect.
Open to full-time roles and collaborations in climate, AI, and fashion tech.
Roles I am looking for: Data Analyst, Analytics Engineer, Data and AI Engineer.