Dmitry Petrov

1.3K posts

Dmitry Petrov banner
Dmitry Petrov

Dmitry Petrov

@FullStackML

🛠️ Building data infra for AI/ML. Ex-Data Scientist @Microsoft. Created DVC, now DataChain. PhD in CS. Serious about data. Less serious about everything else.

San Francisco, CA Katılım Ekim 2011
548 Takip Edilen2.3K Takipçiler
Dmitry Petrov
Dmitry Petrov@FullStackML·
Going deeper this Thursday - with a live Claude Code demo over raw Physical AI data in S3. 8am San Francisco / 11am NY / 4pm London / 5pm CET luma.com/3jjb56bx
English
0
0
0
102
Dmitry Petrov
Dmitry Petrov@FullStackML·
Strongly agree. Today's coding RLVR mostly targets functional correctness, not design or maintainability. The real leap comes when RLVR adds multi-objective evaluation for design, code quality, and maintainability. Then comes the sequel, @dexhorthy: why great software still needs humans 😄
English
0
0
1
110
dex
dex@dexhorthy·
gave up on waiting on nikita for the articles fix - part 1 is here part 2 is coming x.com/i/article/2078…
English
81
214
1.8K
527.5K
Dmitry Petrov
Dmitry Petrov@FullStackML·
@signulll My ex-teammate made the same choice. A US PhD and Big Tech experience made it much easier for him to start a company in China.
English
0
0
0
2K
signüll
signüll@signulll·
kimi’s founder & ceo got his phd at cmu. why didn’t he stay in america?!
signüll tweet media
English
1.2K
415
8.9K
7.6M
Dmitry Petrov
Dmitry Petrov@FullStackML·
Coding-agent intuition breaks down on physical-world data That’s my AI Engineer World’s Fair Online Track talk this week. See you there! I’ll be the guy in the Villain-Con shirt 😄
Dmitry Petrov tweet media
English
1
1
4
154
Dmitry Petrov
Dmitry Petrov@FullStackML·
Claude and ChatGPT are both down. Tokenmaxing seems to work.
Dmitry Petrov tweet mediaDmitry Petrov tweet media
English
2
0
4
638
Alexy 🤍💙🤍
Alexy 🤍💙🤍@ChiefScientist·
This is how we vibecoded in the 1990s
Alexy 🤍💙🤍 tweet media
English
2
0
10
339
Dmitry Petrov
Dmitry Petrov@FullStackML·
@fantopy It’s about Data Agents over S3. Semantic Layer for Files.
English
0
0
0
12
Fantopy
Fantopy@fantopy·
@FullStackML ooh nice, what's your session about? been meaning to check out agentconf stuff
English
1
0
0
46
Dmitry Petrov
Dmitry Petrov@FullStackML·
Talked about data agents beyond SQL at @pydatalondon today. Files, video, sensors, metadata, Python workflows - all the messy stuff where real data work happens. Thanks everyone who came by 🙌
Dmitry Petrov tweet media
English
0
0
5
166
Dmitry Petrov
Dmitry Petrov@FullStackML·
@trikcode 5 —> 128 Another proof that 10x productivity isn’t a limit
English
0
0
0
127
Wise
Wise@trikcode·
Before AI, I had 5 unfinished projects. After AI, I have 128 unfinished projects.
English
316
825
8.1K
209.1K
Dmitry Petrov
Dmitry Petrov@FullStackML·
Excited to speak at @PyData London this Sunday! I’ll be talking about data context for agents beyond SQL: video, sensors, files, metadata, and Python workflows. 10:15 · Hardwick Hub Conference Room Come say hi!
PyData@PyData

Tomorrow, PyData London 2026 begins. If you have your ticket, we'll see you there. If you don't have yours yet — there's still time: hubs.la/Q04jpJ3R0 - - - PyData London 2026: Convene Sancroft, St. Paul's Tutorials: Friday, June 5-7 Talks: Saturday and Sunday, June 6-7 Keynote Speakers all 3 days Tickets: hubs.la/Q04jpJ3R0

English
0
1
2
372
Dmitry Petrov
Dmitry Petrov@FullStackML·
@Hans365days Depends on the problem but yes, context engineering usually takes majority of the time. In data likely 90%
English
1
0
0
37
Hans
Hans@Hans365days·
@FullStackML I have said it before. Choosing a model is only 5% of the job. The rest is engineering, testing and optimising context and workflow. In fact, context engineering is 90% of the job.
English
2
0
1
58
Dmitry Petrov
Dmitry Petrov@FullStackML·
@alexanderbenz Yes, data lineage plays an interesting role in data agents. With human it's nice to have, with agent is absolutely must-have.
English
1
0
1
87
Alexander Benz
Alexander Benz@alexanderbenz·
@FullStackML The shared lesson is that data agents need a trust boundary before they need a prettier chat surface. The hard part is making lineage and refusal visible to the person asking.
English
1
0
0
115
Dmitry Petrov
Dmitry Petrov@FullStackML·
@echpochmac8 Math with trits -1, 0, 1 is cleaner. Physics becomes a mess. My academic grand-supervisor worked on this. The fact that he moved on says a lot.
English
1
0
5
1.1K
simply ri
simply ri@echpochmac8·
Пример корневой ошибки: однажды IT индустрия из-за нескольких лентяев отказалась от троичного кода в пользу двоичного и на этом решении выстроила всю инфраструктуру и ПО. А теперь им не хватает памяти и энергии для вычислений.
Русский
71
2
313
58.2K