Телеграм канал «Nik в мире данных»

Сходил на пати людей искусства.🖼

Ну, думаю, сейчас узнаю, как качественно делать постпродакшн в кино, писать сценарии, оценивать импрессионизм и сколько нужно спецсредств для вдохновения 🍿

Узнал, какие модели с открытыми весами используют в постпродакшне, кто лучше помогает писать сценарии - Claude Code или Codex, и как дела у Higgsfield в КЗ 📈

Один из плюсов коливинга - такие пати иногда проходят прямо в гостиной, и commute занимает 15 секунд (как и commute обратно 😂).

P.S. Пати огонь. Как же всё-таки социализация зависит от места проживания 😘

Вижу активный переход на dashboard-as-a-code, у нас, например, ныне сделать дашборд это вайбкодинг на typescript на внутреннем фреймворке, работает гораздо приятнее и лучше всяких Tableau, Looker, ... (мы теперь full stack engineers in data 😂)

Я на этом же фреймворке еще и динамические презентации в стиле byte byte go делаю =)

Вот и Claude запустил Live Dashboard + Live Animations - https://claude.com/resources/articles/dashboards-and-motion

Главное, харнессинг настроить 🍷

Скам интервью запросов в линкедине.

Словил, видимо, бота, рассылающего вредоносные запросы в линке. Выглядит топорно, но на всякий случай пошарю для тех, кто в активном поиске

Hi Mr. Nik,

I hope you’re doing well!

We’re building an exciting Web3 onchain ecosystem and are looking to bring in senior technical experts, particularly around data engineering, data platforms, analytics, and AI/data infrastructure.

<бла-бла-бла спаршенный опыт в линкедин>

We’re planning a 9–15 month engagement with a $2.2M–$2.5M overall budget, and we’re looking for a small team of 6–8 senior engineers and technical specialists.

Would you be open to discussing a full-time or part-time Data/AI Engineering or technical leadership role on the project?

После дали чуть больше деталей

Hi Mr. Nik , glad to hear from you!

I wanted to share some more focused details on the project and where I believe your experience could be particularly relevant.

We’re currently developing an on-chain Web3 ecosystem around Metaverse game tokens and NFT assets. As the platform evolves, we’re looking for experienced data engineering and platform expertise to help review the existing architecture, strengthen the data infrastructure, and define a scalable technical foundation for the next phase.

The technical scope includes designing and improving data pipelines around on-chain activity, developing reliable data ingestion and processing workflows, establishing data models and lineage, and building monitoring and data-quality processes. We also expect to work with large volumes of blockchain and platform data, so areas such as near-real-time processing, analytics infrastructure, observability, and efficient data architecture will become increasingly important.

We initially worked with another team, but their technical depth and implementation quality did not meet our expectations. We’re therefore looking for experienced engineers who can take a deeper look at what has already been built, identify what should be retained or reworked, and help establish a solid technical foundation for the next phase.

<бла-бла-бла LLM prompt про мой опыт>

Rather than going too deep over messages, would you be open to a conversation with our Tech Manager? He can walk you through the existing architecture, codebase, current implementation, and roadmap, and you can assess where your expertise could have the most impact.

If that sounds relevant, I can send you his calendar link and you can choose a convenient time.

Looking forward to hearing your thoughts!

И финальный message собственно со скамом

<ссылка на calendly>

In the meantime, could you please review the landing page and main logic of current project and prepare your feedback for the meeting?

Here is the latest demo version of the project developed by the previous team, where you can check the tech stack, landing page and staking logic.

Latest Demo 🎮
<ссылка на плохой проект>


Собственно, что плохого, если запустить npm install, то фактически запустится цепочка

The install step runs code automatically. The package.json has a postinstall hook set to npm run start, so just typing npm install launches the server. Real projects never do this.

A hidden download URL is committed in the repo. The file server/config/.config.env holds three base64 values. Decoded, they are a URL at <нехороший сайт>, the header name x-secret-key, and a secret value. That is the classic second-stage loader pattern: fetch an encrypted payload from a throwaway free host and run it.

Дополнительные ред флаги
The "Web3 project" is a repurposed e-commerce template. The server folder is an Indian online-shop boilerplate with Amazon product photos, Paytm payment code, a MacBook image, and sample invoices. A horse-NFT landing page was bolted on top.

The repo has no history. One commit titled "add files via upload", project created on 2 October 2026, five days before you got the link. The group has no avatar, description, or other projects.

The recruiter script is the documented playbook. Unsolicited Web3 or metaverse pitch, flattery copied from your profile, a huge budget, "the previous team disappointed us", a Calendly link, and "please review the demo before the call". Every one of those appears in the published write-ups of this campaign.