Md Majid Jahangir

Implemented Lamport’s distributed mutex algorithm, Basic and Multi-Paxos and evaluated the correctness using DistAlgo.

Performed data analysis and modelling to create a model predicting NYC taxi fare using random forest under the guidance of Prof. Steven Skiena. Used NumPy, Sci-kit, Pandas toolkits and Python. Ranked in top 45% on Kaggle with RMSE of 3.54.

- Implemented an advanced method of authentication (a randomized authentication where a random pair of values are sent to authentication server unlike sending a fixed hashed password).
- Implemented a complete back-end solution in JAVA and a client side application in JavaScript to demonstrate the use-case.

- Several third party applications deployed over Google Assistant or Amazon Echo platform needs authentication of some form or the other. Integrating NetIQ Access Manager (NAM) with such voice assistants through OAuth 2.0 protocol authorizing the third party apps on behalf of the identity protected by NAM.
- Implemented a payments use case, where we developed a third party app, an NLP agent to understand common payments intents such as "check balance", "transfer money to XYZ".
- Deployed… Show more

- Several third party applications deployed over Google Assistant or Amazon Echo platform needs authentication of some form or the other. Integrating NetIQ Access Manager (NAM) with such voice assistants through OAuth 2.0 protocol authorizing the third party apps on behalf of the identity protected by NAM.
- Implemented a payments use case, where we developed a third party app, an NLP agent to understand common payments intents such as "check balance", "transfer money to XYZ".
- Deployed it on Google Assistant Engine and made some code changes at NAM as well as some configuration both at Google and NAM side to initiate seamless authentication over OAuth 2.0 protocol through third party Identity Providers (IdP) like NAM. Show less

- Developed a fully functional utility android application for the Hackfest of our Identity and Access Management (IAM) BU to provide a platform for people to post ideas for the Hackfest, to receive real time push notifications about the event, to receive real time puzzle questions from server and giving members an interface to post answers to the puzzle.

- Implemented a use-case of UMA (User Managed Access Protocol) in health care sector where an identity in one domain wants to access the resource owned by a user in other domain with the user's permission via an android application.
- Implemented a back-end server in JAVA to complete the UMA workflow, for a request to access a patient's medical data by a hospital staff (e.g. a doctor), and send a notification to an android application which provides an interface to the patient to either… Show more

- Implemented a use-case of UMA (User Managed Access Protocol) in health care sector where an identity in one domain wants to access the resource owned by a user in other domain with the user's permission via an android application.
- Implemented a back-end server in JAVA to complete the UMA workflow, for a request to access a patient's medical data by a hospital staff (e.g. a doctor), and send a notification to an android application which provides an interface to the patient to either allow or deny that request. Show less

Using Hyperledger Fabric to develop an interoperable, privacy-preserving data-sharing and exchange decentralized system. Using NodeJS, docker and Go to develop smart contracts and deploy the blockchain network.

An interactive visualization dashboard that visualizes the trends observed in different categories of crimes that occur in United States using techniques like Principal Component Analysis, etc. Implementing the dashboard using HTML, CSS, JS, D3 JS, Bootstrap and Flask.

Traditional UNIX file systems does not provide versioning/backup of files or a feature like "recycle bin/trash" where deleted files reside; because of this there is no way to recover previous versions of file or recover an inadvertently deleted file. bkpfs is such a stackable file system based on Linux 4.20.6 kernel which automatically takes backups of regular files based on policy applied during mount and also provides a user level program to manage such versions/backups like list, view… Show more

Traditional UNIX file systems does not provide versioning/backup of files or a feature like "recycle bin/trash" where deleted files reside; because of this there is no way to recover previous versions of file or recover an inadvertently deleted file. bkpfs is such a stackable file system based on Linux 4.20.6 kernel which automatically takes backups of regular files based on policy applied during mount and also provides a user level program to manage such versions/backups like list, view, delete or restore a specific version/backup of a file. The policy allowed to backup is maintaining N number of versions for each file at a time and backing up files when certain bytes are modified in that file. These options are reconfigurable during every mount/remount of device. Show less

Implemented a loadable Linux kernel module in vanilla 4.20.6 Linux which when loaded into kernel will be supporting a new system call called sys_cpenc. The system call would encrypt or decrypt (using AES cipher in Chaining Mode) or copy a file to another path based on the flag passed. Also, implemented a test program called xhw1 and a dozen of test shell scripts that would receive arguments from user via a command line and would call the system call on behalf of the user and test each… Show more

Implemented a loadable Linux kernel module in vanilla 4.20.6 Linux which when loaded into kernel will be supporting a new system call called sys_cpenc. The system call would encrypt or decrypt (using AES cipher in Chaining Mode) or copy a file to another path based on the flag passed. Also, implemented a test program called xhw1 and a dozen of test shell scripts that would receive arguments from user via a command line and would call the system call on behalf of the user and test each functionality and failures respectively. Show less

Implementing several state based and operation based conflict-free replicated data types in the category of counters, sets and graphs in DistAlgo and Python. Simulating a non-byzantine working model where replicas in a network get updated independently and concurrently without coordination between them and a replica may fail and recover with its memory intact and converge to the same data eventually. Also, testing their correctness criteria and performance analysis and comparing their… Show more

Implementing several state based and operation based conflict-free replicated data types in the category of counters, sets and graphs in DistAlgo and Python. Simulating a non-byzantine working model where replicas in a network get updated independently and concurrently without coordination between them and a replica may fail and recover with its memory intact and converge to the same data eventually. Also, testing their correctness criteria and performance analysis and comparing their performance with https://github.com/neurodrone/crdt in Go. Show less

Performing data exploratory analysis and implementing a light GBM model to predict revenue per customer of Google merchandise store as part of a Kaggle contest using NumPy, Sci-kit, Pandas and Python. Performed permutation tests to see which features boost the model and applied logistic regression to find the top 10 most probable buyers in the future. Ranked in top 40% with RMSE of 1.43.

- Worked as a full time Software Developer, in a team of 2, in designing and implementing the first
analytics solution of NetIQ/Micro Focus, a highly scalable, high availabilty real-time event driven
Analytics Dashboard from scratch, end to end.
- Scaled to handle ~1600 events per second at real-time and ~30 million events from offline storage in a span of 1 hour approx.
- Implemented an asynchronous and push based syslog receiver (both over TCP and UDP) to fetch all
… Show more

- Worked as a full time Software Developer, in a team of 2, in designing and implementing the first
analytics solution of NetIQ/Micro Focus, a highly scalable, high availabilty real-time event driven
Analytics Dashboard from scratch, end to end.
- Scaled to handle ~1600 events per second at real-time and ~30 million events from offline storage in a span of 1 hour approx.
- Implemented an asynchronous and push based syslog receiver (both over TCP and UDP) to fetch all
audit events from NetIQ Access Manager & NetIQ Sentinel.
- Performed asynchronous computations and aggregations on the consolidated data-set (comprising
both real-time and offline data) to push to elasticsearch and project the visualization of the
aggregated data in the customized Kibana dashboard.
- Implemented an asynchronous event-driven elasticsearch client leveraging RxJava in functional programming paradigm. Show less

Worked on Identity and Access Management unit designing, developing and supporting several large scale components. Some of the key implementations in the recent pasts are revamping OAuth 2.0 protocol stack, analytics solution - NetIQ Analytics Server and integration of NAM with HPE Vertica. I also have work experience on multi-factor authentication, various authentication methods and risk based adaptive authentication rule based engine. I have also lead and driven innovation in this domain… Show more

Worked on Identity and Access Management unit designing, developing and supporting several large scale components. Some of the key implementations in the recent pasts are revamping OAuth 2.0 protocol stack, analytics solution - NetIQ Analytics Server and integration of NAM with HPE Vertica. I also have work experience on multi-factor authentication, various authentication methods and risk based adaptive authentication rule based engine. I have also lead and driven innovation in this domain through various channels. Some of the key innovation efforts are incorporating machine learning (ML) in the product during authentication, implementing UBA solutions using ML and data analytics, devising new mechanisms of authentication, etc.
Products/Components handled:
- NetIQ Analytics Server - An analytics solution to aggregate several business parameters and visualize them as charts or stats.
- Identity Provider: A centralized authentication and federation server. Worked on various authentication and federation protocols including SAML 2.0, WS-Fed, OAuth 2.0, OpenID Connect, etc.
- OAuth 2.0 Protocol Stack/ Authorization Server - Revamped the entire stack to make it extensible to support other formats of tokens.
- Risk Based Authentication Engine - Worked and developed some features of the engine to assess the risk posed by user's authentication and adapt the actions based on the risk.
- NAM Collector - Principal Owner of this component . Implemented in JavaScript, it was used to map the event fields of NAM to NetIQ Sentinel event fields. Show less

- Designed and developed an end to end UBA use-case in Java using Apache Kafka, SQL and k-Means clustering algorithm on user’s authentication patterns to classify authentications as malicious or benevolent and take an action based on that using risk based authentication (RBA) and machine learning. Leveraged HPE Vertica and NAM to complete the prototype.
- RBA strengthens the security by assessing the risk of a user using geo-location, cookie context, etc. This use-case strengthens the… Show more

- Designed and developed an end to end UBA use-case in Java using Apache Kafka, SQL and k-Means clustering algorithm on user’s authentication patterns to classify authentications as malicious or benevolent and take an action based on that using risk based authentication (RBA) and machine learning. Leveraged HPE Vertica and NAM to complete the prototype.
- RBA strengthens the security by assessing the risk of a user using geo-location, cookie context, etc. This use-case strengthens the existing RBA framework by using users’ behavioral context and try to assess the user’s risk.
- Used Apache Kafka as a message bus between NAM and Vertica. Implemented a custom Kafka source connector and streaming applications in Java. Show less

- Understanding, designing and implementing OAuth 2.0 protocol and OpenID stack for NetIQ Access Manager using JSON Web Token (JWT) format.
- Also added new functionalities such as access token and refresh token revocation, generation and consumption of JWT via the four basic grant flows; SAML2 bearer profile for OAuth 2.0, backward compatibility for the existing binary format of tokens.
- Improved minor performances issues and the rate of generation and consumption of tokens by 5-8%.

- An attempt to implement a second factor authentication by leveraging the keystrokes metadata of users trying to authenticate, classifying them as real user or a fraud. Implemented an ML model and a back-end server using JAVA and R to train the model of keystroke metadata of several users using Mahalanobis distance metric to cluster data.

- A B2C start-up initiative to provide end customers with salon and spa booking features and management of their appointments.
- Involved as a Co-Founder and tech developer in designing and developing the android application UI, its client side logic and the back-end server to mange such bookings and appointments.
- Implemented the UI of the app on native Android platform and the client side business logic of the application.
- Involved in developing some features of back-end server… Show more

- A B2C start-up initiative to provide end customers with salon and spa booking features and management of their appointments.
- Involved as a Co-Founder and tech developer in designing and developing the android application UI, its client side logic and the back-end server to mange such bookings and appointments.
- Implemented the UI of the app on native Android platform and the client side business logic of the application.
- Involved in developing some features of back-end server implemented in Go Lang.
- Involved in managing the management or the business aspects where I had to interact with salons' managers to strike a partnership deal with them. We successfully had partnership with more than 75 salons/spas across Bengaluru with an average booking of 10-15 per day and a revenue of around 20000 INR or more each month. Show less

- A real-time stock market trading simulation game implemented on native Android platform to provide learning experience to beginners in stock market trading before delving into the real trading.
- Created an interactive UI for android platform which would fetch real-time data of NSE stocks and allow users to place a long/short orders on them, and to provide with notifications based on alerts that are set.

Modified the database structure (implemented using MS Access) to implement differential cut-off system for respective courses and implemented automatic stacked-bar chart generation in PNG format for the data in database using Java. This tool was being used by the Computer Science department of my college.