Bogdan Habic

Nisanka won second place at Nis Hackathon, Belgrade Serbia.

Nisanka is a smart driving companion which gives drivers insightful information regarding their driving habits by connecting to the OBD II port of their car via Bluetooth. Besides giving drivers actionable insights, the application also gives NIS a way to push notifications to driver mobile phones by utilizing contextual information.

The demo itself consisted of emptying the car of its fuel to 20%, and getting a push notification about where the closest NIS gas station qas. After filling the tank, the user gets an in-app reward because the OBD II reader detected a change in the tank itself.

Tenderly is a public procurement platform built on top of Hyperledger Fabric. The platform is unique because participants keep their anonymity by sending only signatures of their offers, so there is no way to connect them to their offer. When the procurement deadline is reached, participants reveal their offers for the first time. This solution removes the need for trust in the procurement process, which is a huge problem in larger companies and some countries.

The goal of this project was to help both Alzheimer patients and their caretakers with regular daily tasks. To do this, we developed four smart home solutions:
1. Smart medicine box which uses a light sensor to check if the patient took their medicine.
2. Bluetooth beacons were used for triangulating the position of the patient inside their home so they could be reminded to bring their keys when exiting their house, or to notify the caretaker if the patient hasn't moved for a while.
3. Alexa push notifications, which were used to notify the patient if they forgot their medicine, keys, or other essential items.
4. A powerful notification system that used the Azure Cloud AI platform and IoT Hub to notify both the patient and the caretaker via SMS, mobile push notifications, and Alexa push notifications if an important message needed to be relayed.

ServiceMap is an internal tool that shows dependencies between microservices. Its value comes out of three features:

1. Inter-service communication which can be whitelisted through ServiceMap itself.
2. Once inter-service communication is whitelisted, the dependency between two services is shown in a GUI.
3. If a service goes down, it is instantly reflected in ServiceMap. This is important because just by following inter-service communication failures you can almost instantly see the root cause of an outage.

RAF Hackathon is a School of Computing organized event that lasted 24 hours. Our solution, Spark, won second place in the competition.

It is a social platform that displays a heatmap of nightlife activity for a city in a given moment. A real-time stream of Twitter and Instagram impressions is processed, extracting time, and location data from them. Using these data points, a heatmap is generated over a city map indicating the most active points of interest. As another dimension to the platform, heatmaps are preserved for previous intervals, so entire nights, days, or months can be replayed to show the movement of activity across the town, as well as activity patterns correlated to the time of day.

AgroCulture is an ad-hoc crowdsourced marketplace for farmers to come together and buy resources they need at cheaper prices due to the economy of scale.

We developed a marketplace for small and medium-sized enterprises to help facilitate cooperation with larger companies based on private blockchain technology.

We used Hyperledger Fabric to create a global network that allows smaller businesses to get noticed and work with larger companies, and for those companies to have clear expectations about the quality of the businesses they are working with.

In order to do that, we designed a metric similar to what a Credit Score is to the crediting industry. We call that metric Business Score. We extrapolate it from various metrics such as the business's ability to follow schedules, produce value up to specification, financial activity of the business, and many other types of data.

Tx Flow allows Ethereum developers to see the whole execution flow for a given transaction easily.

As smart contracts are getting more complex, we are slowly losing visibility of what is happening within a single transaction. This is especially true when a smart contract interacts with other smart contracts.

Tx Flow is what a debugger would look like if all of it had to fit on a single page. The execution is laid out and indented as methods get called, and in turn, call other methods, and if you want to explore a specific point in execution, you can click on it, and all the variables relevant to that line will show up.