A production localization pipeline built entirely on the OpenAI API: one English source of truth, SEO-aware translation prompts, cross-model verification with
All tags
Available
Iurii Uses: #OpenAI
2 projects · 5 articles · 2 reviews
OpenAI's API provides access to GPT models for text generation, embeddings, and function calling. I use it to build AI-powered features like smart search, content generation, and chatbots.
Projects
RAG system on OpenAI and Upstash Vector with 30 language support. Includes streaming chatbot, hybrid search (semantic + keyword), AI ticket classifier, and
Databases
Articles
· 8 min
Semantic search for internal docs with Postgres and pgvector: chunk, embed, and query by meaning so people find the right document, not the exact words.
Stack
Databases
Services
· 24 min
OpenAI API cost optimization: where the money actually goes in production, prompt caching, model routing, budget controls, and observability that surfaces
· 14 min
AI document processing in production: full PDF pipeline — OCR fallbacks, structured output, validation, cost at scale. Beyond naive GPT calls.
· 11 min
Add AI to an existing product without a rebuild. Three integration patterns, how to pick the right one, and what production-ready AI actually demands.
Stack
Libraries
Databases
Services
· 9 min
RAG chatbot for e-commerce: resolve 70% of support queries across 25 languages — ingestion pipeline, hybrid search, confidence thresholds, and streaming UI.
Stack
Libraries
Databases
Reviews
“
We wanted to add an AI feature that turns messy user notes into structured records, but our first attempt returned unpredictable JSON that broke the app half the time. Iurii rebuilt it using structured outputs against the OpenAI API with proper validation, so the data is always shaped the way our database expects. He also added a fallback path for when the model is unsure instead of letting it guess. It's been running in production for a month with no manual cleanup.
Bram de Vries 🇳🇱
Product Lead
Service: ai integration
“
We process thousands of supplier invoices a month and the team was keying them in by hand. Iurii built a pipeline that reads each document, extracts the fields we care about, and routes anything that doesn't match a known template to a human — with a confidence threshold so nothing gets posted automatically that the model wasn't sure about. He was blunt up front that we shouldn't try to automate the low-confidence cases, and that honesty is exactly why finance trusts the system. It's handled around 85% of the volume for two months now with no bad postings we've had to reverse.
Lena Brandt 🇩🇪
Head of Product
Service: ai integration
AvailableGot a project in mind?
Whether it's a new product, an integration, or scaling an existing system — let's talk about how I can help.