About Me
Hello, I’m Aniket, a Software Engineer 3 at Walmart, where I play a pivotal role in the Core Identity and Access Management (IAM) team. At Walmart, I contribute to the design and development of robust authentication and authorization solutions, including single sign-on (SSO) products, used across the entire Walmart ecosystem, including Sam’s Club. Our multi-tenant, highly scalable, and reliable platform supports diverse user groups such as customers, last-mile delivery drivers, marketplace sellers, seller center users, and third-party clients.
My work focuses on building secure, efficient, and resilient systems to meet the complex and dynamic requirements of one of the world’s largest retail networks serving multiple geographical locations like US, Mexico, Canada, China, UK and many more. Before joining Walmart, I honed my skills at JP Morgan Chase as part of their IAM team. There, I developed a multi-tenant system for authenticating and authorizing users in the travel booking domain, ensuring seamless and secure access for a global user base.
I am passionate about creating clean, efficient, and maintainable code that enhances system reliability and delivers a seamless user experience. My technical expertise spans a wide array of programming languages and frameworks, with a strong focus on Java, Spring Boot, DotNet Core, C#, Python, git, and AWS services.
I am most skilled in Software Development, Machine Learning and Watching Movies.
Projects
Reimplementation of the paper Attention Is All You Need in PyTorch
github.com/aniket414/vaswani-et-al-2017
Python, PyTorch, Transformers.
In this project, we re-implement the Transformer model, which was suggested by Vaswani et al., in “Attention Is All You Need” for English to German translation. Our objective is to investigate the Transformer architecture’s efficiency in handling language translation jobs, as well as to find any shortcomings and possible areas for development. We train our model using the Multi30k dataset, which is based on WMT 2016, and test it using the BLEU score as a metric for translation quality Papineni et al., 2002. We efficiently train the model on 4 NVIDIA A100-SXM-80GB GPUs with the help of parallelization in our implementation, which makes use of the nn.DataParallel function from the torch library, see PyTorch, 2023. Surprisingly, despite Vaswani et al.’s efforts to enhance convergence by using residual and layer normalization, we discovered that the Transformer model does not converge as anticipated. This discovery calls into question the Transformer architecture’s robustness and emphasizes the need for additional study and development.
Python, BERT, NLP.
A natural language question answering system that utilizes dialectal understanding in 5 languages 24 dialects to process multiple languages, providing accurate answers based on the question asked.
Nextgen IAM
C#, Python, AWS, Cassandra, Redis.
I developed Identity-as-a-Service APIs to authenticate and authorize both individuals and applications. The service supported various authentication and authorization flows, including OpenID Connect, SAML, Authorization Code, Password, Client Credentials, Authorization Code with PKCE, and Refresh Token.
To enhance analytics, I created a serverless AWS Lambda function that ran daily to generate user data reports. For data streaming, I leveraged AWS-managed Kafka to ensure efficient processing. Additionally, I conducted threat modeling and worked closely with the penetration testing team, ensuring that any security vulnerabilities were promptly addressed.
Nextgen Profile
C#, GraphQL, Cassandra, AWS, Redis.
I developed a Profile-as-a-Service using GraphQL to manage individuals’ profile data efficiently. With GraphQL, users could read, write, update, and delete only the specific sections of their profile they needed, optimizing performance and flexibility. I also ensured compliance with data privacy regulations like CCPA (US) and GDPR (EU) by implementing robust encryption for PII data. This service handled profile operations for millions of Chase Travel customers worldwide, ensuring secure and seamless profile management at scale.
Python, Tensorflow, Flask, HTML.
I started this project as my final year undergraduate project to gain hands-on experience with Machine Learning. I trained a model using the LSTM RNN algorithm to predict solar power generation. The project features an interactive UI where users can select a city, and in the backend, future weather data for that city is fetched via a REST API. This data is then processed by the trained model to predict solar power generation, providing users with insightful forecasts.
Experience
Powering the next great retail disruption.
At Walmart, I work on developing authentication, authorization, and single sign-on (SSO) products that are used across the entire Walmart ecosystem, including Sam’s Club. My role involves building and maintaining a multi-tenant, highly scalable, and reliable IAM platform that supports a wide range of use cases—whether it’s customer logins, last-mile delivery drivers, marketplace sellers, Seller Center, or third-party clients.
Given Walmart’s global scale, I design solutions that cater to its unique identity and access management needs while ensuring the platform remains reliable and scalable enough to handle millions of users daily across various applications and services. I collaborate with cross-functional teams to implement secure and efficient IAM solutions that align with Walmart’s business objectives, ensuring a seamless and secure experience for all users.
If you can't explain it simply, you don't understand it well enough.
I worked as a teaching assistant for undergrad programming course IT-106: Introduction to IT Problem Solving Using Computer Programming and IT-206: Object Oriented Techniques for IT Problem Solving at GMU! The course offers a broad introduction to problem solving and object oriented in IT with emphasis on ideas central to Java.
My major work involved delivering interactive classroom labs and assignments, teaching best code practices and OOP’s concept resulting in a increase in student engagement and their improvement in test scores.
Powering the next great retail disruption.
Walmart operate more than 10,500 stores in 24 countries and employ over 2.3 million associates worldwide. Working directly out of the home office, I found motivation in my daily commutes from hotel through an incredible work environment. As an intern, I worked on the BPMN orchestrator tool Camunda and developed a process instance migration feature which was available only in the enterprise version and not in open source version thus saving a lot of amount. I researched how migration takes place in enterprise version and then designed and developed GraphQL API’s in spring-boot to support the same. To make this even more developer friendly, I also worked on developing a multi-step widget user interface in React.
Alongside Camunda, I also worked on an in house virtual machine monitoring tool and designed API’s in nodeJS to add new feature to our monitoring platform.
I’m confident that I made an impact on the team and found value in the work that I did. As I look back on my past eleven weeks, I cannot believe how much I have grown, technically and professionally. I feel blessed to work with a bunch of smartest Software Engineers, learn from them and work to lead the way in ecommerce universe. And yes, I am no longer exhausted after spending 8 hours of my day sitting in one position. I must say I’m getting much better at writing code day by day.
5G multi-access mobile edge server based security and learning for smart building.
Developed a uniform identity and cryptographic system that enhances security for the Internet of Things devices and cyber-physical systems that service large buildings. Worked on writing entirety of source code and test modules while reading relevant papers and conferring with faculty experts to ensure the software is well designed.
I worked on this project under Dr. Duminda Wijesekera and Dr. Nirup Menon. The project received funding from the Virginia Innovation Partnership Authority under Commonwealth Cyber Initiative for the project: “5G Multi-Access Mobile Edge Server Based Security and Learning for Smart Building.”
Modern buildings strive to enhance resilience and meet environmental goals such as net-zero energy and carbon emissions. Existing building systems use disparate networks, e.g., BACnet, LonTalk, ZigBee, Bluetooth, WIFI, IP to connect with HVAC, elevators, lighting, and physical security systems. These independently controlled systems use proprietary protocols and cloud-based services. Two major cybersecurity problems that arise from this are the lack of overall security governance localized to the building operators’ infrastructure and the inability to communicate between systems securely. Consequently, this affects the ability to optimize the cyber-physical security posture and energy usage by opening a uniform, secure platform to incorporate newer technologies such as solar rooftops and vehicle-to-grid and grid-to-vehicle charging based on price fluctuations during the time of day.
Solving 21st century banking problems through softwares and making game changing products.
Every company needs its softwares appropriately crafted, and JP Morgan Chase is no exception. Developing softwares that is clean, optimized, follows coding best practices is hard, and I played my part in making sure the whole world banking problem is not a problem anymore.
I worked here as a part of a scrum team consisting of 4 Developers, 1 Tester, and a PO/BA. I spent most of my time working on a specific area of an extensive application that our team takes ownership over. The front end is an Angular app that talks to multiple APIs built using a collection of DotNet Core microservices. I spent time working in numerous parts of the stack, including creating APIs, integrating microservices, deploying code, and following best practices in a microservice.
Recently we took the lead on rebuilding the login journey to our application. We, as a team, worked together with the architects and designers to create something that would be fit for purpose and be easy to maintain for years to come. We then built this to a high standard leading to a seamless switch over from the previous system with no issues.
At JPMC, I gained experience working with a continuous integration workflow using tools such as Jenkins, image hosting using Docker, microservice orchestration through Kubernetes, and working with Amazon Web Services for deployments. We built most of our software using stack C#, DotNet Core, Cassandra, GraphQL, and Redis.
For 12 months, I was working on delivering a standalone Identity and Access Management + Profile application that services all our user base. I also gained a lot of experience with Elasticsearch while working on Kibana to provide search functionality for logging. During the project, I spent hours communicating with the onshore team, performing deployments, and giving support.
Towards the end of my time at JPMC, my role involved assisting with the company’s DevOps, including managing our AWS infrastructure and deploying and migrating data. I also spent time configuring and maintaining our applications SSO. This mainly involves configuring new applications and clients who are using ADFS, SAML, Client Credentials and also maintaining our softwares SSO functionality.
I enjoy learning about all areas of our application which allows me to be a knowledge seeker within my team and the other teams in the office. I am always looking to gain improvements and challenge what and how we are working to deliver the best service to the customers while also developing to the highest standard products.
Education
Established in 1957, George Mason University is a public research university. It has seen the best and brightest individuals from across the world pass through its doors, a progressive, and diverse university.
I pursued my Master of Science in Computer Science at George Mason University as a full-time student, specializing in Machine Learning. My time at GMU was an enriching experience that not only deepened my technical expertise but also helped me grow personally and professionally.
During my coursework, I explored various domains of computer science, focusing on machine learning and software engineering. Some of the key courses I took include: Machine Learning, Deep Learning, Natural Language Processing, Artificial Intelligence, Software Development and Software Design Architecture.
- Studied supervised and unsupervised learning techniques, optimization algorithms, and deep learning models.
- Gained hands-on experience in text processing, sentiment analysis, language modeling, and transformer-based architectures.
- Worked with search algorithms, decision trees, reinforcement learning, and AI-driven applications.
- Learned best practices for designing scalable, maintainable software systems using design patterns and architectural principles.
- Focused on agile methodologies, DevOps practices, and full-stack development.
My first year at GMU was full of learning experiences beyond academics. I got better at balancing coursework with my personal life, mastering time management and discipline. Managing finances and budgeting as a student taught me valuable life skills, and collaborating on group projects strengthened my teamwork and communication abilities.
Beyond academics, I embraced the diverse cultural environment at GMU, which exposed me to new perspectives and ideas. I also discovered a few kitchen hacks that helped me significantly cut down my cooking time—an essential skill for a busy student!
Established in 1949, Savitribai Phule Pune University houses 46 academic departments. It is popularly known as the 'Oxford of the East', and is widely regarded in terms of its influence, reputation, and academic pedigree as a leading university in not just the India but also the world.
During my degree at Pune University, I learned many essential skills such as teamwork and project management. Although we mainly focused on software development and the programming principles, I use many of these skills to my advantage in my day-to-day work.
While at University, I served as Technical Secretary, managed labs, students requirements, and ran weekly meetings to oversee the progress of the task in the department. I also organized a department level inter-college technical festival during this time, which was a huge success.
I also had a wonderful experience working as a Campus Ambassador for the Transform Maharashtra’s initiative. Promoted the event and encouraged students to share their ideas on how to tackle major problems being faced by the State Government of Maharashtra. Alongside promotion, I also managed the end-to-end interaction of participants with the organizers, effectively resolving their queries.
Papers
Exploring Machine Learning Algorithms for Gene Function Prediction in Crops
DOI 10.1002/9781394209965.ch6
Jakhmola‐Mani, R., Sonali, N., Pandey, A., Raturi, D., Singh, R., Vanam, K., D, M., Chauhan, R., Katare, D. P., Nongdam, P., & Potshangbam, A. M.
S. Rathod, A. Pandey, M. Pandey, D. Jha, H. Singh
A. Pandey, M. Pandey, D. Jha, H. Singh, S. Rathod
When I'm Not Developing
When I’m not at my desk, I am very active and spend most of my time outdoors (mostly mountains). I do a lot of fitness and recreation activities such as trekking, running, meditation, and trying to make time for things like cooking and watching movies. Since my time at university, I have developed a very keen interest in theatre, films, and play of all genres.
I’m interested in art and music as I like to be creative, this is mainly through films, but I also enjoy abstract art too.
Finally, during the rest of my free time, alongside my interests in software engineering, you will find me researching on:
- Physics
- How to get into space
Look at this wonderful image of Spiti Valley from one of my trips to the majestic Himalayas.
