Skip to content
View ksolano220's full-sized avatar

Block or report ksolano220

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
ksolano220/README.md

Katherine Solano

Marketplace operator · strategy & analytics · applied AI

getzona.app · LinkedIn

I founded and operated ZONA, a two-sided consumer marketplace, owning pricing, venue partnerships, retention, and growth end to end. I work in SQL and Python to find what's actually happening in a business, then ship the fix -- a pricing change, an operating process, or, when it's the right tool, an AI system.

Marketplace Experience

ZONA -- Founder & Marketplace Operator · Oct 2023-Present

Two-sided consumer marketplace connecting nightlife venues (clubs & bars) and consumers. Founded and operated the business end to end across strategy, pricing, partnerships, growth, product, and operations.

  • 150K+ downloads · ~35K MAU · 60+ venue partners · 25%+ week-4 retention
  • Designed the transaction/monetization model -- tiered pricing, deposits, commission -- plus a per-venue partner-ads product, facilitating ~$1M in transaction volume and $200K+ revenue
  • Cohort analysis surfaced a ~4% DAU retention rate as the real growth bottleneck, not acquisition -- reallocated the team around it, then shipped a rules-based decision engine for personalized push notifications that lifted DAU ~15%
  • Ran weekly venue-operator interviews to find reservation-workflow friction and turn it into next-sprint scope
  • Directed the AI voice concierge rollout (Twilio, Claude, Deepgram, Cartesia) automating guest call handling 24/7, live across three venue partners

Selected Work

Marketplace Pricing & Promotion Simulator A contribution-margin decision simulator built on 824K real transaction line items to stress-test take rate, fulfillment cost, promotional investment, and demand assumptions -- including a promo breakeven calculator and a lever sensitivity ranking, so the tool identifies which economic lever moves marketplace profitability the most, not just what happens when you move a slider. Python, pandas, Streamlit, Plotly.

Sentra Autonomous AI agents can take real actions -- sending messages, approving records -- before anyone checks if they should. A runtime policy layer that scores every proposed agent action and blocks, allows, or escalates it before it executes, with cumulative risk tracking and a three-strike shutdown. Demonstrates: governance, risk-based decision rules, systems thinking. Python.

Nonprofit Financial Health Dashboard Is this organization's financial position getting stronger or weaker, and where's the risk? Real IRS Form 990 filings for 20 nonprofits, EIN-verified against known values at fetch time, with illustrative (adjustable) screening thresholds and peer comparisons that suppress themselves when a sector's sample is too small to mean anything. Demonstrates: financial-statement analysis, risk screening, data-integrity discipline. Python, pandas, Streamlit, Plotly.

Healthcare Access Risk Analysis Where should limited access-to-care resources go first? A composite risk-scoring model across county-level population, insurance, and hospital-density inputs, ranking and mapping 3,200+ counties into priority tiers. Demonstrates: geographic prioritization, weighted decision modeling. Python.

Medicare Billing Analysis Hospital pricing for the same procedure can vary 3x+ by location. SQL analysis of CMS Medicare billing data (145K+ records, 3,000+ hospitals) to find cost outliers, markup patterns, and geographic pricing disparities. Demonstrates: SQL, pricing-variance analysis, geographic economics. SQL.

More applied-AI and healthcare analytics projects →

How I Work

Business problem → data → hypothesis → intervention → measurement → decision.

I use code when it's the fastest way to test the hypothesis or operationalize the answer.

Tools

Analytics: SQL · Python (pandas, NumPy) · Tableau · Mixpanel · PostHog Product / Experimentation: A/B testing · cohort & retention analysis · unit economics · Streamlit · Figma · Jira Applied AI: Claude API · Hugging Face / PyTorch (LoRA fine-tuning) · multi-agent systems · prompt caching & evals

Background

Founder & operator, ZONA (2023-present). MBA, Pepperdine Graziadio Business School. M.S. Management and Analytics, NYU School of Professional Studies -- Candidate.

Pinned Loading

  1. marketplace-pricing-promo-simulator marketplace-pricing-promo-simulator Public

    Interactive contribution-margin simulator for marketplace pricing/promo decisions, built on real order-level transaction data (UCI Online Retail II).

    Python

  2. sentra sentra Public

    Runtime execution control for AI agents. Policy enforcement, risk tracking, and three-strike shutdown, with a Python SDK for drop-in integration.

    Python 1

  3. nonprofit-financial-health-dashboard nonprofit-financial-health-dashboard Public

    Financial health dashboard for 20 real U.S. nonprofits, built on live IRS Form 990 filings via ProPublica's Nonprofit Explorer API.

    Python

  4. healthcare-access-risk-analysis healthcare-access-risk-analysis Public

    Models healthcare access as a constrained decision system by combining insurance, infrastructure, and population data. Identifies high-risk regions and examines how resource limitations shape patie…

    Python

  5. medicare-billing-analysis medicare-billing-analysis Public

    SQL analysis of CMS Medicare inpatient billing data: identifying cost outliers, markup patterns, and geographic disparities across 3,000+ US hospitals

    Python

  6. care-gap-engine care-gap-engine Public

    Population health outreach prioritization. Ranks open care gaps by (clinical urgency × response likelihood × equity priority), then drafts personalized outreach with Claude.

    Python