Sometimes the problem is mine. Sometimes it belongs to a team, a community, or a user. I find the friction, decide whether it's actually worth solving, build something real, and learn from what happens next.
SightlineDive destination intelligence · I love diving. I don't love researching a dive trip across twenty tabs.
After 115+ dives, I was still piecing together wildlife seasons, conditions, operator pages, and forum advice just to answer a simple question: where should I dive, and when? So I built Sightline, a source-backed way to compare destinations by species, season, conditions, and experience level. I kept the scope narrow, verified sources only instead of crowdsourced trip reports, so every recommendation could actually be trusted.
36 destinations · 79 species · Source-backed
Know where your Express Entry score actually stands
Canada publishes years of Express Entry draw data, but figuring out which draws actually matter to you is surprisingly painful. CRS Compass turns official IRCC history into a personalized view of your score, showing how it compares with the programs and categories you're eligible for and flagging relevant draws you might otherwise miss.
10+years of draw history
Product hiring, monitored at the source
Job boards kept surfacing the same roles while thousands of company career pages changed underneath them. I built an automated pipeline that monitors a 12,500+ company universe across four ATS platforms, discovers and normalizes product roles, filters the noise, and selectively uses LLMs to rank the jobs worth inspecting.
12,500+companies · 4 ATS platforms
Paid engagements with real stakeholders and measured outcomes.
One front door for a program running on 8+ tools
Students and staff were navigating a program spread across disconnected systems. I worked with stakeholders to consolidate the experience into a single BruinLearn front door, with a SmartSuite operations layer behind it for the team running the program.
$2Kpaid · 85% fewer clicks · 5 hrs/week back
Workload-based staffing for a 63,000-student district
The district needed a defensible way to decide where limited specialist capacity should go across schools with very different needs. I worked with district stakeholders to build a workload-based staffing model for nurses, psychologists, and speech-language pathologists that translated service demand into practical staffing recommendations.
$10Kengagement · Nurses, SLPs & psychologists
Volunteering built around people, not listings
Finding a volunteer opportunity is easy. Getting a group of people to actually do it together is harder. Tidepool organizes opportunities around teams, clubs, and friends so people can discover something worthwhile and sign up together.
Tech news without the information overload
Keeping up with technology started taking more time to filter than to actually read. Daybreak automatically collects the day's coverage, cuts duplicate and low-signal stories, and turns it into one short morning briefing.
10 storiesDelivered daily
From job description to application-ready
Applying well meant repeating the same work: deciding whether a role was worth pursuing, checking every requirement, and tailoring the same experience again and again. Tailorbot turns a job description into a structured fit decision and, when the role clears the bar, generates tailored application materials grounded in my actual experience.
2+ hourssaved daily
Meeting research without the tab archaeology
Preparing for a conversation meant reconstructing context across company pages, search results, notes, and profiles every single time. Dossier takes a person, company, and meeting context and turns the research into a one-page prep brief.
1 hour → 1 minute
