ConfirmIQ

Business Data Analytics | Second Year Group Project

Saarah Aqeelah

Rashini Gomes

Poornima Dhananjani

Sri Lanka’s economy runs on commitments — and loses millions when they break. Tourism welcomed a record 2.36 million visitors in 2025, yet revenue per visitor fell, with cancellations and no-shows leaking income hotels never recover. The same failure repeats across every sector: private hospital channelling platforms lose scarce specialist hours to missed appointments; e-commerce sellers on cash-on-delivery — the dominant payment method locally and regionally — pay shipping twice when 20–30% of parcels are refused at the doorstep; and banks and finance companies, recovering from the highest non-performing loan ratio in South Asia (11.97% in 2022), still detect missed installments only after the damage is done. ConfirmIQ is a Commitment Risk Engine built to stop this leakage. Powered by a decision-tree machine learning model trained on real reservation data, it scores every booking, appointment, COD order, or loan installment from 0–100 at the moment it is made, then acts automatically: personalized reminders for medium-risk customers, deposits or prepayment prompts for high-risk ones, and dynamic overbooking or slot release to resell predicted losses before they expire. The model is explainable enough for banks and hospitals to audit, light enough to run on local infrastructure, and retrains on each client’s own data, improving continuously. One engine protects hotels, hospitals, couriers, lenders, and restaurants alike — built in Colombo, priced for Sri Lankan businesses, and architected to scale across South Asia’s commitment-driven economy.