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
Built a Dockerized Python/Ollama monitoring system for configurable web sources and multiple users. The pipeline fetches and cleans pages, discovers new items with an LLM, extracts schema-validated article data, applies a separate processing prompt, deduplicates results, and exports to CSV or Telegram. Includes async jobs, retries/exponential backoff, concurrency controls, SSRF/redirect/response-size protection, persistent state, admin/user controls, automated tests, and CPU/GPU Compose profiles.
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
Challenges
View all
Related posts
The AI Clothing Inspection System is an advanced automated quality control solution for garment and textile inspection. It leverages cutting-edge artificial intelligence technologies including multilingual OCR, deep learning vision models, and real-time object detection to ensure comprehensive product quality assurance. The system automates the detection of defects, verifies compliance with order specifications, and generates detailed inspection reports.
spotting garment defects with vision plus OCR on labels is a solid real-world use
Your support queue is growing faster than your team can handle.
What if you could resolve 60% of support tickets automatically with AI?
I'm building Relay AI, an AI support agent trained on your docs, tickets, and product. It answers customers in seconds, and escalates to your team only when it should.
Exceptional as always.
AI Lease & Contract Intelligence for Property Managers is a local web application that converts rental agreements into structured, searchable data. Users upload PDF or TXT leases, and the system extracts tenant, property, rent, security deposit, lease dates, notice periods and renewal terms, clearly marking any missing or uncertain details rather than inventing values.
The platform includes a RAG-based contract assistant that answers questions with cited source passages, using a local Ollama model when available and keyword retrieval as a fallback. Automated review flags surface unclear clauses and missing information for human verification, while an expiry tracker and live dashboard help managers stay ahead of renewals and expirations.
Built with Python, Flask and SQLite, it runs entirely on the user’s machine with no paid APIs or subscriptions, keeping sensitive documents private.
I noticed the system flags missing lease fields instead of guessing values, which should cut down on manual verification headaches.