🔄

I'm currently updating my portfolio. Check back soon for new work!

Selected Works

A focus on complex data systems, AI-driven automation, and Fintech innovation. Turning operational complexity into intuitive, scalable experiences.

V i e w
Mydoh Earn Mydoh Save Mydoh Receive Mydoh Money Movement Mydoh Earn clone Mydoh Save clone Mydoh Receive clone Mydoh Money Movement clone

Mydoh

What it is: A family finance app serving 300K+ parents and youth, helping families teach kids money management.

Target users: Parents and their children (typically ages 8–18)

Value:
  • •Increased activated users by 200%, driving more funded accounts and completed onboarding
  • •Boosted account funding rates, leading to higher transaction volume early in the lifecycle
  • •Strengthened parent-child engagement, increasing ongoing usage and retention
Fintech Mobile App
V i e w
Lease Portfolio Dashboard

Lease AI

What it is: An AI-powered tool that reads and structures commercial lease agreements into clear, usable data.

Target users: Commercial real estate operators, landlords, brokers, and tenants

Value:
  • •Turned complex leases into structured, digestible insights
  • •Reduced manual review time
  • •Improved clarity around key lease terms and obligations
AI SaaS Workflow Automation 0-1 Web
V i e w
Eqwitty
Search documents, investors, or legal filings...

Series A Data Room

Legal & IP

8 files

Financials

12 files

Cap Table

4 files

Organized: Legal & IP

Smart AI auto-sorted 2 files

Progress

0% 33% 66% 100%

Legal & IP Verification

Financial Statements

Cap Table Management

Eqwitty

What it is: A platform that streamlines the end-to-end investment process for founders and investors.

Target users: Startup founders, angel investors, and venture capital firms

Value:
  • •Simplified fragmented workflows into one platform
  • •Improved transparency and trust
  • •Enabled faster, more efficient fundraising
0-1 AI Fintech Deal Management
V i e w
app.invoiceai.io/upload

Invoices

CS

Drag document here

Portfolio Services

Bill to:

1847 Restaurant Row
San Francisco, CA, 94105

Invoice details:

Invoice date 05/03/23
Invoice # RS-9942
Due date 05/31/23
Description Rate Amount
Leasing AI $2,500.00 $2,500.00
AP Automation $3,200.00 $3,200.00
Newsly App $1,800.00 $1,800.00
Eqwitty Platform $4,500.00 $4,500.00
Mydoh Mobile $2,100.00 $2,100.00
Subtotal:$500.00
Sales tax (8%):$40.00
Total due: $14,100.00

Invoice Review

#RS-9942 • May 03, 2023

AI-Extracted Line Items

01
Leasing AI
$2.5k
02
AP Automation
$3.2k
03
Newsly App
$1.8k
04
Eqwitty
$4.5k
05
Mydoh
$2.1k
06
Yello
$1.5k

Invoicing

What it is: A digital invoicing and financial workflow tool for businesses.

Target users: Small business owners, freelancers, and operators

Value:
  • •Reduced administrative overhead
  • •Improved cash flow visibility
  • •Simplified invoicing into intuitive workflows
AI Fintech Automation 0-1 Web
V i e w

YellO

What it is: A mobile health passport that enables patients to book vaccinations, track medical records, and securely own their health data.

Target users: Patients navigating fragmented healthcare systems, especially those managing vaccinations.

Value:
  • •100+ user sign-ups during early MVP rollout
  • •Defined a 0→1 product strategy across patient and provider experiences
  • •Established a foundation for secure data ownership and sharing
Health Blockchain 0-1 Mobile App
V i e w

Model

N=24 CANDIDATES
Lakeshore Towers
$2.4M
Park Avenue Plaza
$1.8M
Riverside Heights
$3.2M
Central Station Hub
$2.1M
Skyline Residences
$1.5M

Zurich

What it is: A portfolio optimization model that uses data and constraint-based logic (ILP) to prioritize public assets.

Target users: Public sector asset managers and urban planners.

Value:
  • •Maximizing impact and capital efficiency across large-scale real estate portfolios
  • •Data-driven prioritization for public investment
  • •Scalable framework for managing complex asset constraints
Data Science Optimization Real Estate Public Sector
Health Blockchain Mobile App

Winner of Blockhack Global 2020 for best business proposal!

YellO: A Decentralized Health Passport for Patients

Role

Co-founder, Product Design Lead

Focus

Product Strategy & 0→1 Design

📱 Product

A mobile health passport that enables patients to book vaccinations, track their medical records, and securely own their health data.

🎯 Who it’s for

Patients navigating fragmented healthcare systems, especially those managing vaccinations and records across multiple providers.

The Bet

"If we give patients ownership of their health data and simplify key actions like booking and record sharing, we can reduce friction and improve trust and engagement in healthcare interactions."

⚠️ Constraints

  • •Fragmented ecosystem with poor interoperability
  • •High privacy and security requirements
  • •Limited funding requiring a focused MVP
  • •Balancing complexity with simple UX

📊 Outcomes

  • 100+ user sign-ups during early MVP rollout

  • Defined a 0→1 product strategy across patient and provider experiences

  • Designed core flows for booking, check-in, and record tracking

  • Established a foundation for secure data ownership and sharing

✂️ What we cut

  • •Expanding beyond vaccinations too early
  • •Complex medical record systems
  • •Feature-heavy experiences that slowed development

Overview & The Problem

Healthcare systems are difficult to navigate and lack a unified patient experience. Patients often rely on disconnected providers, manual processes, and outdated tools to manage their health information.

The core issue was systemic fragmentation: health data is siloed, patients lack direct access, and paper-based systems introduce errors and inefficiencies.

Insight

The problem was not just access to data, but control and trust. Patients needed ownership over their health records, confidence that their data is accurate, and a simple way to interact with complex systems.

The Solution

We designed YellO as a simple, patient-centered experience built around three key actions:

Book appointments

Seamless vaccination scheduling

Check-in instantly

Digital check-in at the clinic

Access records

Verified portable health data

Design Highlights

Simplified a complex system

Reduced a fragmented healthcare journey into three clear actions.

Designed for trust

Focused on transparency, validation, and user control over data.

End-to-end thinking

Connected patient and provider experiences into one system.

Built for adoption

Focused on vaccinations as a high-frequency entry point.

Key Features

Digital Health Passport

What it does

  • •Verified vaccination records in scannable format
  • •Share records instantly with institutions
  • •Single source of truth across systems

Why it matters

Replaces fragile paper records like the physical Yellow Card and builds trust through tamper-resistant data.

Seamless Booking and Check-in

What it does

  • •Book appointments in seconds
  • •Reminders and confirmations
  • •Instant clinic check-in

Why it matters

Reduces administrative burden for everyone and eliminates redundant manual paperwork.

Secure Data Ownership and Sharing

What it does

  • •Records are securely stored and validated
  • •Users choose when and with whom to share
  • •Consistent data across all providers

Why it matters

Shifts control from institutions back to patients and improves overall trust in the healthcare system.

Why It Worked

  • •Narrow scope enabled faster execution
  • •Clear mental model improved usability
  • •Addressed both user and system-level problems

Takeaway

I helped design a 0→1 healthcare product that simplifies how patients access and manage their health data by focusing on ownership, trust, and usability.

Research News Enterprise

Newsly: Turning Fragmented Research into Actionable Insights

Role

Product Design, UX/UI, and Front-End Support

Focus

End-to-End Research Platform

📱 Product

A personalized news and sentiment platform that helps enterprise teams discover, analyze, and share insights faster.

🎯 Who it’s for

Analysts, investors, and research teams who rely on timely, high-quality information to support decision-making.

The Bet

"If we centralize news, personalize it by user interests, and layer in sentiment analysis, we can reduce research time and enable faster, more confident decision-making."

⚠️ Constraints

  • •Large volume of data across fragmented sources
  • •Need to balance powerful functionality with simplicity
  • •Limited existing infrastructure for personalization and search

📊 Outcomes

  • Reduced research time by 20% through centralized discovery

  • Defined a scalable structure for topic-based personalization

  • Improved ability to discover, organize, and share research

  • Created a foundation for faster collaboration and decision-making

✂️ What we cut

  • •Overly complex analytics dashboards
  • •Static, non-customizable news feeds
  • •Feature-heavy experiences that increased cognitive load

Overview & The Problem

Enterprise research workflows are heavily dependent on news, but existing tools make it difficult to efficiently find, organize, and interpret information.

Users relied on multiple platforms and subscriptions. Research was time-intensive, often taking 1 to 3+ hours daily. Articles were difficult to store, retrieve, and organize.

Insight

The problem was not access to information; it was the ability to process and act on it efficiently. Users needed a centralized place for discovery and faster ways to interpret content.

The Solution

We designed Newsly as a centralized, personalized research platform that brings together discovery, analysis, and collaboration.

Follow Topics

Custom feeds based on user interest

Analyze Sentiment

At-a-glance industry perception

Save & Share

Structured knowledge organization

Design Highlights

Centralized experience

Replaced fragmented workflows with a single, unified interface.

Personalized discovery

Enabled users to create and follow topics tailored to their needs.

Reduced cognitive load

Introduced sentiment indicators to speed up understanding.

Flexible organization

Allowed users to save, group, and revisit content easily.

Key Features

Topic-Based Feeds

What it does

  • •Curates content based on user-defined interests
  • •Eliminates the need to search multiple platforms
  • •Keeps users updated in real time

Why it matters

Increases relevance of content and reduces time spent searching for high-signal news.

Sentiment Analysis Layer

What it does

  • •Classifies articles as positive, neutral, or negative
  • •Provides at-a-glance context for market trends

Why it matters

Helps users quickly identify signals and identify trends without reading every article in full.

Collections and Sharing

What it does

  • •Stores articles in a structured, accessible way
  • •Enables collaboration across teams

Why it matters

Improves knowledge sharing and makes research reusable across the organization.

Why It Worked

  • •Centralization reduced switching costs
  • •Personalization increased signal-to-noise
  • •Sentiment layer sped up interpretation

Takeaway

I transformed a fragmented research workflow into a centralized, intelligent system that helps enterprise teams move from information overload to actionable insights.

Data Science Optimization Real Estate

Project Zurich · Portfolio Optimization

Role

Product Designer and Data Scientist — Led the design of the system end to end, including the front-end experience and the optimization model powering decision-making.

🏢 Context

Infrastructure Ontario manages a massive and diverse real estate portfolio of over 40 million rentable square feet and 1 million acres of land used for various provincial programs.

📉 The Problem

The current funding model is insufficient to optimally maintain and operate every individual asset within its massive and diverse portfolio, which exceeds 40 million rentable square feet and 1 million acres.

Reactive and Asset-Specific Process

The current method for categorizing assets as Core or Transition is largely reactive. It depends on external triggers from clients, such as a ministry declaring a program surplus or a building’s condition deteriorating to a point where the Facility Condition Index is greater than or equal to 20 percent, rather than a proactive evaluation of the portfolio’s overall value.

Lack of Holistic Oversight

Because the process is driven by individual asset triggers, Infrastructure Ontario lacks a comprehensive portfolio-wide view. This makes it difficult to identify strategic opportunities to allocate limited funds and resources across more than 10 different asset types.

Strategic Alignment Challenges

There is a need to ensure that the selection of Core assets aligns with broader provincial strategies and social objectives while adhering to strict capital and operational budget constraints.

Approach

📋 The Asset Performance Scorecard

To enable a proactive approach, defined a data-driven scorecard to benchmark each asset’s performance:

  • Condition: Facility Condition Index and Deferred Maintenance

  • Financials: Funding Gaps, Operating Expenses, and O&M demands

  • Utilization: Occupancy levels

  • Strategy: Capital Renewal Program status, Ministry Projects, and Asset Types

⚙️ The Methodology: Integer Linear Programming

Applied Integer Linear Programming to determine the best mix of assets that maximize portfolio value under strict budget constraints.

✅ Solution

Designed a portfolio optimization system that evaluates all assets collectively and recommends which to retain as Core, monitor, or transition. The system shifts decision-making from reactive triggers to proactive, data-driven evaluation and enables alignment with financial, operational, and policy constraints across the full portfolio.

Core Logic · Optimization Model (ILP)

Objective Function

max Z =
n ∑ i=1
Xi · ACVi

Subject to:

n ∑ i=1
Xi · Costi ≤ BudgetTotal
Xi ∈ {0, 1}

Methodology Context

Integer Linear Programming is a widely used global optimization technique for solving constrained selection problems. A strong precedent is the Finnish government’s Forest and Biodiversity Program, which applied ILP to maximize conservation value using defined metrics and constraints:

Metric Strategic Constraint
Total Purchase Cost ≤ $215 million
Endangered Species ≥ 1,000 species
Old Broad-Leaved Trees ≥ 3,000 m²
Natural Water Economy ≥ 22 (minimum threshold)
Total Conservation Area ≥ 75 hectares
Closest Natural Reserve ≤ 200 km

These constraints illustrate how ILP balances multiple performance variables while enforcing strict real-world limits.

Impact

Holistic Oversight

Provided a comprehensive view across more than 10 asset types simultaneously.

Strategic Alignment

Ensured asset selection aligns with provincial strategies and social objectives.

Optimized Resources

Enabled proactive prioritization of high-value assets within annual capital and operational constraints.

Portfolio Evolution

Shifted asset management from reactive categorization to portfolio-level optimization.

Maximizing Portfolio Impact Through Data-Driven Optimization

Designed a portfolio optimization system using Integer Linear Programming to help Infrastructure Ontario prioritize public assets, shifting from reactive decision-making to data-driven capital allocation across a 40M+ sq ft portfolio.