rohan bhansali

Yes. That Deloitte.

Exactly the one you are thinking of.

The Data Sciences team at Dentsu is a product team building Dentsu Marketing Cloud (DMC), an ecosystem of data-driven tools to create, analyse and optimise digital ad campaigns.

• Led an initiative to provide greater integration between all DMC products, allowing DMC to be sold to prospective clients as a whole suite (instead of subscriptions to individual tools).
• Redesigned several legacy tools from scratch to use modern APIs and best practices, with a major emphasis on reducing user actions (and subsequent confusion) as well as decreasing page load times by over 60%.
• Redesigned all input forms on the website using best UX practices to improve clarity and brevity. use ML based suggestions to predict user behaviour
• Worked closely with the back-end team to build a new performance monitoring tool powered by Google Analytics.
• Conducted a site-wide comprehensive CSS / JS Audit to reduce page load times and improve site performance by over 70%

Expensify is a leading fintech company automating receipt transcription, expense report submission and management for companies of all sizes worldwide.

• Built processes to monitor and remove expired and/or unused PCI data for GDPR compliance.
• Collaborated to develop a tool to manually confirm automated OCR data from receipts for clients who required internal audits of receipt data.
• Participated in building a system to automate responses for customer support to decrease wait times and queue sizes for known recurring issues.
• Managed internal Chrome extension that optimised workflow of Sales staff to convert leads across various customer acquisition tools and services.
• Contributed to internal tooling to help diagnose and remediate client policy and billing issues.

Zappos is one of the largest online retailers selling shoes, clothes and other accessories. It is subsidiary of Amazon and also has a sister/budget site called 6pm.

• Contributed to the development of a responsive CSS+JS framework to be used across all Zappos mobile and desktop sites.
• Developed the 'My Account' portal for an upcoming responsive redesign of the mobile websites of Zappos.com and its sister site, 6pm.com.
• Designed and built a WebView to be used in the Zappos mobile apps as an interface for a new machine learning based recommendation engine.

An all-in-one ticketing app for attendees and organisers for small-scale events. Sort of like ticketmaster but without all the fees, headaches and setup costs. Designed with a very low barrier for entry - download, signup and create a paid event in less than 5 minutes.

• Create events with custom limits for tickets and different tiers (early bird, VIP etc).
• Browse through events filtered by genre, date, area or price.
• Allow organisers to act as "bouncers" and scan encrypted QR code tickets with instant back-end verification.
• Accept payments from attendees using phone camera and OCR to extract card details. Payments processed through Stripe.
• Allow organisers to cash out payments when the event is complete using PayPal.

A service to track crowd levels in dining areas in Universities, designed to scale outwards to eateries in the local community. Locations crowd-sourced from the app acted as a secondary method of data collection. We had accurate crowd data even before the app was installed on a single user's phone.

• Raspberry Pi sticks used to ping routers and pull number of connected devices to WiFi router(s) in each location.
• Both iOS and Android apps to self-report crowd data using GPS and get real-time information for a desired location.
• Subscribe to notifications at specific locations to get current crowd levels at custom time and date intervals.
• Used a formula of historical data, Raspberry Pi pings and live crowdsourced data to determine the "crowdedness" of a specific place.