Han Insights: Finance Research & AI Automation

Context

Gabriel writes market research for a professional financial newsletter, but you won’t find his name on it. At 15, he doesn’t run his own LinkedIn account, so the analysis he writes goes out under a mentor’s name instead of his.

The newsletter is Han Insights, founded and led by its Founder & Chief Economist, previously an economist at Cboe Global Markets specializing in U.S. equity and derivatives markets, market structure, and exchange economics. Gabriel had already been researching companies for Junior Economic Club pitches; this internship was a chance to do that kind of research for a real, outside professional audience, working alongside other college-age interns rather than only classmates.

My Role

Two things at once: researching markets, macro developments, geopolitical events, commodities, and public companies to write newsletter-style summaries, and building the Python- and AI-assisted tooling the team used to produce the newsletter day to day.

Process

A real research cycle looks like this: scan for a topic actually worth writing about, check the claim against its source, draft it, check it again, and only then does it go out. AI tools help with the first and third steps: scanning for topics and drafting language. Topic selection, the angle taken, and final approval stay human. Nothing is published without someone deciding it’s right, not just that it reads well.

That standard carried into a longer piece of work, too. Separately, and on his own time, not as Han Insights work, Gabriel wrote an independent research project on how AI governance affects value creation in financial services. It’s still being prepared for outside review and isn’t published on this site yet, but it came out of the same instinct he practiced here: verify a claim before it goes in the piece, and cut it if it can’t be confirmed, even if it would have made the argument stronger. The same instinct, applied to a different kind of claim, shows up again in the Algorithmic Trading Research System he built: deciding what a piece of work is not allowed to say.

Result

At least one newsletter-style investment analysis Gabriel wrote was published this way — through Han Insights, via a mentor’s LinkedIn post, his research behind someone else’s byline. The link isn’t posted here yet while the final details get confirmed.

Reflection

Writing about a shipping lane’s risk premium for a newsletter audience is a narrower skill than it sounds like. Ninety seconds, one read, no follow-up questions — that’s the actual constraint, not “explain oil markets” in the abstract. Working under someone who’d already done that professionally, at a market-structure level, meant the bar for what counted as a finished piece was a real one, not a school-assignment one.

None of it carries his byline yet. The newsletter reads the research, not the name attached to it, and that’s fine — the work is what’s being learned from either way.