Hi 👋

I'm Aleksei

London, UK

I'm a full-stack engineer and entrepreneur building end-to-end AI-powered products — from backend and frontend to scalable infrastructure.

Working on now

Anrie @ Unreal Labs

Founding member; joined at day zero. Building an AI coworker for performance marketing teams at a Sequoia and First Round-backed startup.

Maya11:32 AM

@Anrie send me new creative types my competitors are running this week

⏳ 1

AnrieAPP11:32 AM

👀 Found 3 new creative types across 5 competitors.

🎬 UGC demos with objection handling — 3 brands.
🗣️ Talking heads with founder POV — 2 brands.
📸 Before/afters showing workflow pain — 1 brand.
Best angle: founder-led narrative with pain → solution arc.
🧪 I turned the top examples into 3 testable angles for your next pack.

✅ 3🚀 2😎

Leo9:14 AM

@Anrie what new ads are competitors testing in our category?

⏳ 1

AnrieAPP9:14 AM

👀 Competitor scan complete.

🎬 11 new UGC ads launched this week.
👤 4 brands testing founder-led narratives.
📢 3 brands shifting from discount copy to outcome claims.
Best angle: before/after showing workflow pain.
🧪 I added examples and a creative brief for your next pack.

👀 4✅ 2😎

Nina2:08 PM

@Anrie check all Meta ad sets against our $45 CPA target and recommend actions.

⏳ 1

AnrieAPP2:08 PM

✅ Performance scan complete.

Pause: 🔻 Broad Demo Static — CPA $71, CTR down 38%.
Pause: 🔻 Interest Workflow Pain — CPA $64, frequency 4.8.
Scale: 📈 UGC Problem/Solution — CPA $32, 41 purchases.
Scale: 📈 Founder Hook Variant 03 — CPA $36, strongest hold rate.
💸 Recommended move: pause 2 ad sets and shift 20% of budget to the winners.

🎯 4❤️ 2😎

Omar4:26 PM

@Anrie create a new pack from the winning UGC angle and launch it after approval.

⏳ 1

AnrieAPP4:26 PM

🚀 Pack ready: May 12 UGC Scale Pack.

🎨 6 creatives generated.
🧩 3 hooks × 2 formats × 1 CTAs.
🏆 Built from winner: UGC Problem/Solution.
🛡️ Meta drafts prepared with naming, UTMs, and budget guardrails.
✅ Reply “approve” and I’ll launch it.

✅ 3🚀 1😎

Previously

Terra Quantum

Lead AI Engineer; built voice AI agents and led published research on LLM/CV model compression.

Published Research

Literally Me

AI meme face-swap Telegram bot with 150K+ MAU, 1.5M+ users, and ~$70K revenue in 2025.

UnrealMe

Consumer iOS app for personalized AI videos from selfies, built in 2024 before GenAI video models matured. Built with custom FLUX LoRAs for personalized image generation; I owned the full cycle from product development to the ML/AI backend pipeline for training LoRAs.

Literally Me Photo

AI photo generation Telegram Mini App built end-to-end.

Papers and Projects

Whisker / HouseViewer

1st place at Black Forest Labs & NVIDIA London Hackathon; turns floor plans into immersive 3D homes.

whisker houseviewer demo linkedin x

TQCompressor

IEEE MIPR paper on GPT-2 compression; 1.5× smaller with no quality loss and 30× less training data.

paper hugging face github

TetraOpt

Q1 MDPI Applied Sciences paper on CV model optimization; compressed ResNet-18 by 14.5× with minimal accuracy loss.

paper