Software engineer and former Code Fellows instructor.

I'm Adam Owada, based in Seattle. I build Python data systems and applications with AI coding tools. My work ranges from soccer data covering 18 leagues to software for construction teams and textile designers. Teaching hundreds of developers taught me how much a clear explanation matters.

Selected work

Client work · Sports data

Agentify

Built Python data pipelines covering 18 soccer leagues and 10,000+ player profiles, automating collection and reconciliation previously handled in Excel. The owner uses the administration dashboard daily.

Independent project · Trading systems

Trader-trader

I lead the architecture, product direction and testing of this Python trading application. One optimization made an offline full-history computation 5.1 times faster, with identical results.

Co-founder · Construction software

Observe Safety

Building a construction safety application for field reports, photos and corrective actions, including offline report drafts. Currently in alpha testing with my partner.

Creative tools · Textile design

Selvedge Repeat Studio

My sister is a textile designer. Her requirements became a browser app for arranging repeating patterns, saving editable projects and exporting artwork to use alongside Photoshop.

Experience

  1. Observe Safety, LLC

    Co-Founder and Full-Stack Engineer

    Building software for construction safety reporting and follow-up, with web and mobile workflows. Currently in alpha.

  2. Alignerr

    Contract AI Model Evaluator

    Compared coding-agent transcripts and evaluated their behavior against detailed project rubrics.

  3. Contract Full-Stack Developer & Data Engineer

    Built data pipelines, salary-comparison tools and an administration dashboard as one developer on a small team.

  4. Outlier.ai

    AI Model Evaluation Contributor

    Evaluated and rewrote model outputs for coding, function-calling, and tool-use tasks using detailed rubrics and quality criteria.

  5. Code Fellows, Inc.

    Lead Instructor, Python

    Taught hundreds of students in Python and full-stack development, with an 87.2 average Net Promoter Score across seven classes.

Contact

Have a project in mind, a question about my work, or an opportunity to discuss? I'd be glad to hear from you.

Email Adam

Research and writing

I build test harnesses to investigate how coding agents work. My latest published study compared eight configurations across 400 scored implementations, measuring the tradeoffs between quality and token use.

Read the 400-run study
First page thumbnail of the RuleLedger v3 white paper.

June 17, 2026

RuleLedger v3: Measuring Reasoning Effort in LLM Software Engineering

A 200-run study of how reasoning effort affected GPT-5.5 software-engineering quality, including the frequency of weak results.

Open PDF
First page thumbnail of the Spark Mode Efficiency white paper.

June 25, 2026

Spark Mode Efficiency: Direct Edit vs Proposal Mode

A 360-run comparison of coding-agent workflows: direct edits, proposed changes and a single agent, scored for quality and token use.

Open PDF