The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Back to feedPost
Developed AI-powered semantic search systems using Weaviate as the core vector database. Designed hybrid search (keyword + vector) architecture and integrated large language models for contextual retrieval and summarization.
Optimized schema design, embedding pipelines, and indexing strategies to support large-scale document corpora and conversational AI systems.
The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Related posts
I am developing a new AI agent tool that is seed-backed. We are currently building a comprehensive design system, UI components, and test cases, including edge scenarios.
I am solely responsible for designing the entire application, including architecture flows.
Nice 🔥🙌
AI Video Production — Landing Page
Hand-coded landing page for Nureeff AI — an AI-powered video production agency based in Romania and Egypt. Built with pure HTML/CSS/JS from scratch, no builders or templates.
The project required a high-impact editorial design with bold typography, custom cursor, scroll-reveal animations, sticky navigation, and a full FAQ accordion — all implemented in clean vanilla JS without any libraries.
Key deliverables: hero section with stats, services grid, process timeline, formats showcase, features block, FAQ, and a dual-contact CTA for two regional offices. Telegram and WhatsApp integration, Schema markup (Organization, FAQPage, ProfessionalService), full SEO optimization, and EU-compliant privacy policy linked.
PageSpeed 90+. Mobile-optimized. Deployed on free hosting with custom domain.
Built a Multimodal RAG system live on stream. From zero to working CLI in one session.
The goal: a terminal-based RAG pipeline using Haystack, Gemini Multimodal Embeddings, and Flash Lite, querying complex research PDFs through a conversational loop.
Three real engineering lessons from the build:
Chunking isn't optional. Raw PDFs exhaust token limits instantly. 6-page splits solved it.
InMemoryDocumentStore works for prototyping, but re-vectorizes on every launch. Persistent DBs like Weaviate are the next step.
Haystack's pipeline is strict. Mis-wiring a retriever to a generator crashes the loop immediately. API contracts matter.
Result: a CLI that maps user queries to the top 4 semantic chunks in RAM and returns grounded, non-hallucinated answers from your own documents.
Every mistake, every fix, on camera.
"Ugurcan Uzunkaya this is seriously impressive! 🔥 Building a full RAG pipeline live on stream in one session is no joke!
That chunking lesson is so real — token limits hit different when you're working with raw PDFs.
And the Haystack pipeline point about API contracts is...
Trending
aivideo
AI video tools are moving at warp speed. Which ones are you experimenting with?
illustration
Handcrafted illustration is bubbling up across the web. What are you drawing lately?
aidesignflow
AI tools are redefining design work. What's your current workflow?
returntonature
Spring is a reset for creativity. What’s inspiring you outside the screen right now?
freelancerlife
Freelancer life is wins, pivots, and everything in between. What’s yours right now?