Fraud detection
For one of the biggest job portals in Poland (www.poloniusz.pl) I wrote a system that reduces the amount of manual work they put into checking if the offers reported by employers are in line with the law and the portal’s Terms and Condition. The portal on average gets about 200 new offers daily, and checking an offer is a complicated task – so even if they spend only 2 minutes on each offer, it sums up to 200 manhours monthly. With the help of the new system, they were able to reduce the amount of work to 18 manhours monthly – achieving a nice 91% efficiency improvement.
About Me
SUMMARY: ● I specialize in ROI optimization using ML / AI: - segmenting clients + differentiating their offers (personalization) - recommender systems (products with the highest purchase probability) - suggesting traffic acquisition with the best predicted ROI - fraud detection - data auto-correction ● I use a mix of different tools, mostly python + scikit-learn + keras + MySQL
NEW FREELANCER!
