Projects using Ollama in Cairo
Projects using Ollama in Cairo
Sign Up
Post a job
Sign Up
Log In
Filters
2
Projects
People
Message
0
Marwan Mahmoud
Built a self-verifying RAG pipeline that refuses to hallucinate confidently — uses an independent reranker (not the generating LLM) to judge answer quality, requires 4-source consensus, and scales responses across a 5-tier confidence system down to hard refusal.
0
17
Message
0
Yossef Ayman Zedan
LeadBot is a B2B lead automation platform I built that turns raw scraped LinkedIn data into scored, tiered leads with personalized outreach ready to send. It extracts company and profile links from raw page data, runs each lead through a scoring pipeline that checks hiring signals and applicant volume before spending on deeper enrichment, then classifies leads as Cold, Warm, or Hot with transparent reasoning behind every score. For qualified leads, it generates personalized connection notes and follow-up messages based on the specific role or signal that made them worth reaching out to, not generic templates. Built with LangGraph orchestration and Ollama for the scoring and message generation, with every lead persisted so nothing gets processed twice.
0
19
Message
0
Yossef Ayman Zedan
InvoiqAI is an AP invoice automation tool I built that reads invoices, purchase orders, and delivery receipts, cross-checks them against each other, and routes approvals automatically, so AP teams stop manually keying numbers and stop paying invoices nobody actually verified. It runs full three-way matching on every document, flags price or quantity mismatches with plain-English reasoning, and catches duplicate invoices before they get paid twice. Approval rules route small invoices to auto-approve and larger ones to the right person, so nothing sits in a queue waiting on a human unless it actually needs one. Tested across 85+ real documents with a 96.5% extraction pass rate.
0
32
Message
0
Yossef Ayman Zedan
OneRecruiter is an AI-powered candidate screening tool I built for small staffing agencies, where AI screening rejects good candidates and consultants can't afford to look like a black box. Every criterion in a role's rubric is scored independently with a transparent, auditable rationale, so recruiters see exactly why a candidate scored 90% instead of trusting a mystery number. The standout feature is candidate pooling: agencies upload their CV database once and it stays matched against every future job, so a candidate sourced for one role automatically surfaces for the next relevant opening. Validated it end-to-end with a real staffing agency client, running a live CAD Designer search against 12 candidates with synthetic test data before their real screening.
0
62
Explore projects
