Rıdvan Ünlüerler

• Built an AI-powered call analysis system covering sentiment analysis, automated call scoring, summarisation and evaluation of customer interactions
• Designed and developed a React dashboard for tracking agent performance
• Wrote Excel automations that cut down repetitive daily reporting
• Optimised SQL queries running over large call-centre datasets

• Developed and optimised ABAP queries and ALV tables for structured business reporting
• Designed SAP user interfaces and worked with consultants to keep technical documentation aligned with business requirements

A progressive web app that turns a day's visit list into a planned route. Once the stops are in, it works out the order, the arrival times and the estimated finish, then tracks where you are through the day.

• Address search, saved stops and prepared tours for fast entry
• Stop order optimised from a traffic-aware distance and time matrix
• Chained arrival times from departure time, time spent per stop and an optional return to the start
• Map requests routed through a Node.js proxy I wrote, so the API key never reaches the browser
• The active tour lives on the server, so a route left on a phone continues on another device

Vanilla JS, Leaflet, Node.js, TomTom Routing and Search API

Open the site

The paper checklist used in retail audits, rebuilt for a phone. It collects the lists that change with store size onto one screen and turns a finished form into a PDF at the end of the day.

• Checklists that change with the store's size segment, marked off with a single tap
• Device-first: forms can be filled with no connection and sync once one is back
• Edits made on two devices merge by last update, and deletions are tracked separately so they stay deleted
• An admin account can read every user's records
• Completed forms rendered to PDF in the browser

Vanilla JS, Service Worker, PHP, jsPDF

Open the site

Computer vision for underwater footage, where colour loss and scattering break most off-the-shelf detectors.

• CycleGAN used to restore colour and contrast in degraded underwater images
• YOLO run on the enhanced frames to detect marine life
• The two stages combined into one pipeline so detection works on raw footage

Python, CycleGAN, YOLO

A cross-platform mobile app for recording workouts and following progress over time.

• Built once for both platforms with Flutter
• Training history and progress kept per user

Flutter, Dart

• Fourth year. Graduation thesis on underwater computer vision.

• One-week programme on the Verint workforce management platform and shift planning.

• Certified in PADI Emergency First Response, primary and secondary care
• Teaching under pressure is where I learned to keep a clear head and explain things simply

• Led the club and its social responsibility projects for two years
• Handled the club's finances, balance sheets and records before that
• Organised Derin Akıntı VI, the club's symposium