Founding Member & Engineering Lead @ Pice | B2B Payments

Abhisekh Nag  • Peerlist

Continuously changing and improving the APIs both in terms of architecture and features based on rapidly shifting market demands. Involves building new features , maintenance and scalable architecture design of the backend API.
Technologies: PostgreSQL , NestJS , Sequelize , Microservice Architecture

PagarBook Backend Core API

Designed and developed a complex system to maintain a pre-processed cache of dynamic inteligence data for millions of malicious IPs and Domains to minimize fetch time.
Technologies used - Python, ElasticSearch, NSQ, Docker, GraphDB, Gremlin, Java, Redis.

Threat Intelligence Precompute Cache System

This project involved maintenance and enhancement of an Invoice Processing System.
This system took hard copy invoices, scanned them and used OCR to convert them to text and matched them against regular expressions to pick out relevant data and validate them according to client requirements and export them to SAP.
I contributed in the enhancement and training of the OCR system and adding UI features and validations according to user requirements. Also worked on developing a workflow to automate the entire process for a percent of invoices where all data are valid.

Digital Invoice Processing

Developed a mininal network packet sniffer that could capture and display details about packets captured by the network adapter. 
The software when deployed on a PC would try to turn on promiscuous mode on the network adapter and sniff out all the transiting network packets and display them in the UI in real time. 
This was developed in java using javaFX and jpcap library.

Network Packet Sniffer

Design and Development of a complex yet elegant backend system which gathers cyber threat data from numerous sources from across the internet and maintains a distributed database of all that data and also serves that data in a presentable way to the frontend UI.
Technologies Used - Python, JaunusGraph, ElasticSearch, NSQ, Redis, Java, Docker, Cassandra, NodeJS.

Threat Intelligence Backend

A system to detect threats and security issues in a targeted network using cyber threat intel gathered from numerous sources across the internet. This system can detect threats drilled down to each system on the network. It also creates security incidents detailing the threats as well as the remediation for the network owners to respond to.
Technologies used - Python, Elasticsearch, Git, Airflow, Google BigQuerry, PostgreSQL.

Vulnerability Intelligence Backend

Developed a web based student fees management system using J2EE.
created a Web Site which featured access restrictions and hosted a system to input, store, manage and display student fees information for an institution.
The web site featured a Notification system that sent alerts to admins regarding pending fees and also an authorization based access.
The development was done in Netbeans using the Apache Tomcat Web Server.

Student Management System

Developed a system to classify text components from a set of random input images. 
The system would take images as input, apply dynamic binarization on it and then select the connected components from it and pass it through a Support Vector Machine which would classify the components as text or not text according to the features defined.
This was developed in MATLAB.

Optical Character Recognition

Developed a flash game in Flash CS6.
This game featured a 2D character which can be controlled by the user with the arrow keys in the keyboard to dodge bullets coming out of sentry guns placed in the screen.
The user input module was designed to handle multi key inputs and process them logically to translate them to character movement.
The  scripting was done using action script 3.0.
The artwork was developed in Flash editor.

Flash game

Developed android game using the Unity game engine and box2D physics engine. 
This game features a 3D figure character that the user can control using touch/tap on android devices and advance through the levels by dodging obstacles and solving puzzles.
The scripting was done in C# and the artwork was developed in corelDRAW and Photoshop CS6.