Mizuho · Consulting engagement
AI specialist via Synechron. Production data platform, agentic ingestion, stakeholder communication, and delivery ownership.
New York · Finance × AI
Hi, I’m Jay.
Always building.
I turn rough ideas into useful systems—and usually find another question along the way.
Mathematics → investing → AI.
It’s been an interesting route.
Try a real decision
A minute in my shoes.
Pick an approach. I’ll show you the call I made.
Mizuho / Platform launch
The app works. The UI could be better. The schedule is tight.
Business users want an AppKit rebuild. The Streamlit version already supports the workflow. What do you prioritize?
Two reasonable priorities. One actual project.
Selected work
A few things I’ve put into the world.
Open a project to explore its scope, decisions, and what comes next.
Beyond the bullet points
Still curious.
Still building.
I like turning a rough, rambling idea into something I can run, inspect, and improve.
What draws me to AI?
During graduate school, machine learning connected my interest in mathematical patterns to systems that could make predictions. That fascination has stayed with me through investment research, trading, and building applications.
What did Mizuho teach me?
A useful system has to work for the people around it. Listening to business users, coordinating data teams and QA, making trade-offs, and supporting a launch have all changed how I build.
What am I looking for next?
Ownership of direction and delivery. I enjoy leading development and staying close enough to the technology to make sound decisions. I want to learn through more customer problems and contribute those lessons to better products.
What question keeps following me?
How could agents keep useful context and learn through exploration? IRIS is one place I am investigating that question. It is an experiment I want to take further.
The path here
A few different lenses.
Alysida Capital · Cofounder
Quantitative trading and strategy. Hired and trained four researchers and developers, with the team growing to ten including summer interns.
Synovation AI · ML & strategy systems
Financial machine-learning research frameworks and reinforcement-learning environments for decision-making.
Investment research & futures trading
China-focused equity research at Meritz Asset Management in Seoul: SVM screening, company visits, and fundamental reports. Earlier intraday futures trading at Jichengli.
For people who bring an agent
A résumé your tools can read.
A structured profile, a plain-text version, and small Python tools to explore the facts or prepare an interview brief. No account or API key required to run the downloaded kit.
Download the agent kitProfile + two scripts + usage guide · ZIPStructured profileExperience, project scope, and evidence limits · JSONPlain-text profileDrop it into your preferred assistant · MarkdownProfile readerRead or filter the profile locally · PythonInterview brief builderPair a job description with the profile · Python
After unzipping the kit
# Read the project evidence
python3 profile_reader.py --section projects
# Prepare context for your assistant
python3 prepare_interview.py --role role.txtThe tools read local files and print text. They do not call an AI service or rate the candidate.
These are profile tools created for this site. They are separate from the source code for IRIS and my work projects. Read the guide ↗
Let’s compare notes
Have a problem worth building for?
Based in New York. Open to travel and relocation for the right opportunity.