Skip to content
Seraphimodality · ꙮ
Open separately ↗
 shipulin.md

Head of Product Research · Yandex Search & AI / Paris

Aleksei
Shipulin.

CV ↗LinkedIn ↗Projects ↓Get in touch ↗

I lead research for Search & AI at Yandex. I build products, put ideas to the test and write about what I find.

Selected projects

All projects ↗
RAISE Summit 2026 / Team projectTeam hackathon prototype

FlowTwin

FlowTwin: we won the Google DeepMind track

Our team, Low Cortisol, took first place in the Google DeepMind track at RAISE Summit 2026. I helped shape the project, designed with agents and operated the demo.

Read the story : FlowTwin
A screenshot of the FlowTwin demo: hospital floors, synthetic patients and a timeline.
Screenshot of the team’s FlowTwin demo. Open aggregated waiting-time data and synthetic patients; a prototype, not a hospital deployment.
2024–2025 / Research toolReleased tool

Grounded

Grounded: I built a bot for open-ended answers

I built an AI tool that proposes categories, codes responses and explains its choices, so the cost of analysis need not decide which questions researchers ask.

Read the story : Grounded
OPEN-ENDED RESPONSES CATEGORIES A B C
From responses to categories. A conceptual diagram with illustrative text fragments.
2025–2026 / AI and walkingReleased Telegram bot with open source code

Bot Voyage

Bot Voyage: I built a tour guide I walk with

I made a Telegram bot for a workshop in 2025, then went home from a walk to fix it. Location, facts too good to trust, and an interface that can afford to wait.

Read the story : Bot Voyage
WALK, NOTICE, ASK
A walk as a sequence of stories. An illustrative route, not a city map.
2026 / Computational experimentInteractive demo

Fly Poetry

Fly Poetry: I adapted a fly-brain demo for Chukovsky

I replaced the video in badapple-fly with a poem to understand what it means to “run” something on a fly brain. An adaptation of others’ open work; the fly is not reading.

Read the story : Fly Poetry
TEXT CONNECTOME BODY
Letters → stimulation → movement. A conceptual diagram, not reading or simulation output.
A FAMILIAR MOVEMENT WHAT CHANGES IF WE TRY IT ANOTHER WAY?
A familiar knot as an experiment. A drawing, not tying instructions.

From the blog · a personal experiment

64 shoelace observations, then someone explained the knot

In July 2023 I published a personal experiment comparing four ways to tie my laces. Statistics found a difference; a reader’s comment helped explain the mechanism.

Read the essay ↗

Beyond work

A notebook
with side quests.

A tap on a phone that calls Claude, a button missing from NPS, Nabokov in the margins. Notes on things I cannot quite leave alone.

Open the blog ↗

Or try the ꙮ in the header. The notebook has a strange entrance.