Skip to content
View vladyslawwww's full-sized avatar
💘
💘

Highlights

  • Pro

Block or report vladyslawwww

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
vladyslawwww/README.md

hi, i'm vladyslav. ☕️

i'm a 20-year-old software engineer and ai builder based in Poland. i specialize in ripping out slow, manual enterprise workflows and replacing them with native, scalable ai pipelines.

i don't just "plug in" ai. i build the bullet-proof backend logic, the async architecture, and the edge-case handling required to make models survive in production environments.

currently architecting enterprise-grade multimodal extraction engines scaling across 21 countries @ inter cars s.a.


⚙️ what i actually do (and how i do it)

i bridge the gap between heavy enterprise operations and modern ai capabilities.

  • the problem: large companies rely on rigid legacy template-based ocr tools or slow workflows for processing thousands of complex documents, invoices, and tables. they break on edge-cases.
  • my solution: i build custom ocr & vlm (vision-language model) engines from scratch.
  • how it works under the hood: i write asynchronous python pipelines deployed on serverless azure durable functions. i use azure ai foundry for data extraction, but i wrap the ai outputs in strict pydantic schemas and heuristic fallbacks. this guarantees 100% predictable, structured json before the data ever hits the company's erp system.

🛠️ the expanded tech stack

core engineering & cloud infrastructure

ai, data validation & automation

web


🚀 what i'm focused on right now

  • 📉 optimizing cloud costs by routing easy data extraction tasks to deterministic python scripts, saving expensive llm tokens only for complex edge-cases.
  • ⚡ exploring the intersection of heavy backend architecture and enterprise citizen development (power platform).

🌍 the human behind the code

i believe in extreme ownership and adapting fast. at 14, i moved from odessa, ukraine to poland entirely on my own to study. by 19, i broke into the corporate enterprise world. by 20, i landed my software engineering role. i know how to navigate deep waters.

when i'm not writing pydantic schemas or optimizing cloud architectures, you'll probably find me:

  • dialing in the perfect cup of specialty drip coffee. ☕️
  • hanging out with my girlfriend & our cat, puree (she is the real tech lead around here). 🐾

🤝 let's connect

find me talking about system architecture, ai workflows, and handling AI edge-cases here:

linkedin  •  website  •  email

Pinned Loading

  1. undocs-saas undocs-saas Public template

    Lightweight SaaS Boilerplate & Template for AI-driven document extraction. Includes RBAC, Stripe billing, and Gemini Vision integrations.

    HTML