Aleksei Petrov
CTO at QuantFlow; builds AI agents that integrate with CI and issue trackers to automate coding and delivery with telemetry and controls.
TurboQuant Compresses LLM Cache to 3‑4 Bits, No Loss
TurboQuant from Google Research is wild. They basically found a way to compress LLM KV cache and vector search down to ~3–4 bits, with zero accuracy loss and even faster runtime than 32‑bit keys. https://research.google/blog/turboquant-redefining-ai-efficiency-with-extreme-compression/
Anthropic's Upcoming Autonomous Agent Mirrors Pilot
Anthropic’s new tool is on the horizon, pretty close to what the Pilot is - autonomous agent. Let’s see if it’s a good one ✌️
Pilot, ClaudeCode, GLM 5.1 Hit 74.2% Success
Pilot + ClaudeCode + GLM 5.1 74.2% success rate on Terminal Benchmark 2.0 full run. I have to check we didn't violate any rules 😁
Opus Plummets to 10th as Token Costs Rise
Opus dropped on the benchmark from 2nd to 10th place. Limits are tighter, tokens are more expensive now, this is how it all goes. Investment needs to be returned. Nice I don’t care, as my limits are cooked till the end...
GLM Tackles Tough
GLM caught the “chess-best-move” trial on this test run — the nasty one. It’s been grinding on it for ~40 minutes already, ~20 left on the clock. Let’s see if it clears it; Opus only managed to pull it off a...
Reduced to One Worker, Now Runs Overnight
Scaled infra down to a single worker. Last run was burning tokens way too fast. Now it’s crawling… so this one runs overnight.
AI Agent Automates Dev Issue Resolution on AWS
AI Agent passing development issues, online. Join the Thread 🍿 AWS infra. ClaudeCode. Pilot. Hit the star: https://github.com/qf-studio/pilot
Design a Website with One AI
Website Design w/ Claude is so much fun. NO SKILLS NEEDED. Sonnet 4.6 made a quick sketch to refresh our website. Not bad for the single prompt 👏 Screenshot: Product section. Terminal animated.
Solo AI‑Built Pilot Tops Terminal‑Bench 2.0
Pilot — #1 on Terminal-Bench 2.0. 82.9% accuracy. 124 entries. Claude Opus 4.6. Built by single person + AI in Montenegro. No VC. No cluster. Standard infra. Leaderboard is live: https://www.tbench.ai/leaderboard/terminal-bench/2.0 Open source: https://pilot.quantflow.studio
5K/Mo Buys Full AI‑powered Dev Studio
QuantFlow Studio is open for subscriptions 🎉 $5K/mo — one EU dev's cost — buys a whole studio's output. Engineering, design, AI integrations. 2 engineers orchestrating self-made agents, end-to-end. Proof we're not LARPing: • Pilot — 82% on Terminal Bench 2.0 (built in...
14 Releases in One Day, Delivery Fully Automated
14 releases one day. Delivery on autopilot 🛩️ Just checked reports, Claude and Pilot are building.
AI Agents Delivered Fully Tested Code Overnight
Set up two AI agents before bed last night. - Pilot (executor) — picks GitHub issues, writes code, ships. - ClaudeCode (/loop to monitor) — checks status every 30 min, reports. Morning: everything wired, tested, parity checks passing I review with a coffee...
Pilot Ships with Short Video and GIF Demos
Pilot on delivery duty today. Cutting short videos and gifs to show how it ships. https://github.com/qf-studio/pilot
Pilot v2.86.3 Adds Crash Cleanup, Dashboard Graph, Repo Migration
Pilot v2.86.3 released. Fixed: — Stale worktrees after OOM/SIGKILL never cleaned up (818MB each) — Squash merges dropped PR titles → broke release tagging — GoReleaser pointed to old repo after migration New: — Dashboard git graph follows active task's project — Worktree cleanup on crash and...
Top AI Coding Agent Ignored Despite Benchmark Victory
Built an AI agent that took #1 on Terminal-Bench 2.0 — "the industry benchmark for coding agents". 82.0% across 445 trials. Validated by the maintainer 3 days ago. "Ready to merge." Still not on the leaderboard. LinkedIn DM — no response. Discord —...
