CTO · Tokyo ⇄ Düsseldorf

Building things that
matter

Engineer, builder, and lifelong learner. This is my corner of the internet — where professional craft meets personal life.

Curious about me? Ask Ramu — he answers anything about me from what I've shared with him.

Projects

Things I've built

Live, running on Hiren's AI Lab

Reports & Studies

Findings, written up in full

Measured findings from Hiren's AI Lab, written up in full and free to download.

shastrarth · Findings report How five small language models argue 1 October 2026 · v1.3 · 14 pp
Latest · Findings report

How five small language models argue

Findings from 2,500+ instrumented turns across 250 controlled debates and 300+ turns of public debate on Shastrarth

  • Gemma and Qwen argue best (3.39 and 3.32 of 5) and are statistically tied; Llama, Mistral and Phi form two tiers below
  • Model size predicts neither quality nor speed — the 8B Gemma beats the 7.2B Mistral on every judged dimension
  • Llama is the fastest by a wide margin: 25.9 s per turn, against Mistral's 59.3 s

Also in the library: Deep research by a team of four small language models: measured, step by step — see all 2.

Family Portal

A private corner for the people who matter most.

"In family life, love is the oil that eases friction,
the cement that binds closer together."

— Friedrich Nietzsche

The person behind
the screen

Hiren Joshi

I'm a builder at heart — drawn to problems that sit at the intersection of engineering and real-world impact. Whether it's self-hosting services, writing code, or tinkering with infrastructure, I believe the best way to understand something is to build it yourself.

This website is my digital home: part portfolio, part family portal, part ongoing experiment. It runs on a server in my home, and built piece by piece — like everything else I care about.

Let's connect

Whether it's a project idea, a question, or just a hello — my inbox is always open.