Work history

Architected and Designed: Lead the architectural design and implementation of large-scale microservices-based systems, ensuring alignment with business goals and technical requirements.

Technical Problem Solving: Identified and resolved complex technical issues across development, QA, DevOps, and performance testing teams, enhancing overall system stability and performance.

Cross-Functional Leadership: Guided and collaborated with development, QA, DevOps, and performance testing teams to ensure seamless integration and high-quality delivery of microservices.

Client Communication: Acted as the primary technical liaison with clients, providing regular updates on development progress, addressing concerns, and ensuring satisfaction with project deliverables.

Code Review and Quality Assurance: Conducted thorough code reviews, ensuring adherence to best practices, coding standards, and architectural guidelines. Provided constructive feedback and approved new feature deployments.

Performance Optimization: Oversaw and directed performance testing efforts, identifying bottlenecks, and implementing solutions to optimize system performance and scalability.

Continuous Integration/Continuous Deployment (CI/CD): Collaborated with DevOps teams to streamline and automate the CI/CD pipelines, ensuring reliable and efficient deployments of microservices.

Mentorship and Training: Provided technical mentorship to developers, fostering a culture of continuous learning and improvement within the team.

System Integration: Managed the integration of various microservices and third-party systems, ensuring seamless communication and data flow across the entire architecture.

Documentation and Best Practices: Developed and maintained architectural documentation, coding standards, and best practices, ensuring consistency and knowledge sharing across teams.

Risk Management: Identified potential risks in system design and deployment, proposing mitigation strategies to minimize impact on project timelines and system performance.

Stakeholder Engagement: Engaged with stakeholders to gather requirements, provide technical insights, and ensure alignment of project goals with business needs.

Scalability Planning: Planned and executed strategies to scale microservices architecture to accommodate growing user demands and future business needs.

Technologies: ASP.Net Core REST API, Azure Service Bus, Azure API Management, Azure Kubernetes Service (AKS), Entity Framework Core, Redis, MS SQL, Docker, Azure DevOps, Open Telemetry, Azure Monitor, Microsoft Entra ID, Azure Managed Identities, XUnit.Net.

Holberg EEG AS is a Norwegian (European) eHealth company that specializes in advanced diagnostic data interpretation and sharing in the field of Electroencephalography (EEG).

With a suite of software products including SCORE Premium, hiSCORE, autoSCORE and freeSCORE Holberg EEG AS is leading the way in standardized structural interpretation and reporting of EEG data.

As a Senior Technology Lead, I was entrusted with complete ownership of SCORE Premium and worked alongside a talented team to develop hiSCORE, freeSCORE, and autoSCORE. Through this journey, I was privileged to become the Solution Architect for the entire Holberg EEG product suite, working under the guidance of the CTO.

Brief overview of each product:

SCORE Premium is a legacy desktop application that is gradually being phased out, as the company transitions its customers to the more advanced hiSCORE web application. SCORE Premium utilizes plugins to integrate with different EEG review software.

Technologies: WPF, WCF, MS SQL, Telerik Reporting, Entity Framework

hiSCORE are cloud-ready web platforms. It uses a proprietary application layer protocol called "EEG Interchange Protocol" (EEGIP) to integrate with EEG review software. I designed and developed EEGIP, demonstrating exceptional technical prowess and problem-solving abilities. I was responsible for the overall architecture and development of hiSCORE.

Technologies: React, ASP.Net Core REST API, Entity Framework Core, SignalR Core, Redis, RabbitMQ, ZeroMQ, MS SQL, Docker, Kubernetes, Azure DevOps.

autoSCORE is a software-only medical device that uses AI models to categorize normal and abnormal EEG recordings. Holberg EEG has the world’s largest collection of accurately labelled anonymized EEG recordings. Except for AI model training, I had exclusive ownership of the overall architecture and contributed to all other aspects of the product.

 Integrated autoSCORE into the Natus EEG software, featured in a JAMA journal publication, advancing neuroscientific research. (Publication: [https://jamanetwork.com/journals/jamaneurology/fullarticle/2806244]) 

Technologies: TensorFlow, Python, ONNX Runtime, C, C++, Google Test, Boost Libraries, Open SSL. 

Enterprise Resource Management (ERM) is a suite of solutions for project management in capital-intensive engineering industries. It provides integrated resource management at every stage of procurement and construction. The software gives complete control over the project, from beginning to decommissioning, and can be customized to meet customer needs.

The modules of ERM include Material Manager, Catalogue Manager, Planning, Production, and Supplier Portal. For maximum efficiency, ERM can be fully integrated into the customer's process.

I was a Team Lead in the Aveva ERM Material module which is critical to ERM. I was a Lead Developer in the ERM core framework and Integrated Ship Building team responsible for the extensibility and integration of other Aveva and third-party solutions into Aveva ERM.

I contributed to the design and development of the ERM integration and ERM audit log frameworks.

Technologies: WCF, WPF, .Net 4.5, ASP .Net MVC, Xceed WPF Controls, LightInject, Fluent Validation Framework, Oracle 11g, Devart dotConnect for Oracle, NUnit, NLog and Custom ORM Framework with Custom Query Language called LQuery. 

Transport for London (TFL) was using an old ticketing system called FASTIS+ (Flexible and Secure Ticket Issuing System + Oyster). This is a legacy desktop application developed using VC++ and intended to be used only by the clerks who issue tickets in all the Railway Stations across London.

TFL wanted a new SOA based system that will replace the existing legacy system in multiple phases. Cubic as part of its Next City Vision, designed and developed a new Ticketing System known as “Core Ticketing Engine” (CTE) that will replace the existing system.

Cubic designed and developed 10 Engines that collectively form the Core Ticketing Engine. Following are the names of the Engines:

1. Journey Planning Engine (JPE)

2. Ticketing Engine (TE)

3. Fulfilment Engine (FE)

4. Customer Relationship Management Engine (CRM)

5. Peripheral Driver Engine (PDE)

6. ToD (Ticket on Departure) Engine (ToD)

7. Shifts Accounts and Settlements Engine (SASE)

8. ITSO Engine (IE)

9. Data Sync Engine (DSE)

10. Data Distribution Service (DDS)

All the Engines are WCF Services. These services are intended to be used by multiple ticketing channels.

I was involved in the development of PDE, JPE, TE, FE and ITSO Engine. I had the ownership of JPE, TE, FE, CRM, DSE and PDE for FTO and TVM.

Technologies: WCF, .Net 3.5, MySQL, MS SQL, NUnit, NLog, and Entity Framework. 

Infosys Front End Automation Platform: was an incubation project intended to automate ticket resolution for various clients. The platform in the current form is deployed at 3 different client locations. I was involved in the end-to-end design and development of this platform including client proposals.

Technologies: Sikuli, .NET 3.5, Windows Batch Scripting, IBM RFT and Windows Workflow Foundation.

NCBC Back-office Automation: As a part of my onsite consulting assignment, was involved in the development and maintenance of the NCBC (National Commercial Bank Capital, Kingdom of Saudi Arabia) intranet SharePoint 2007 portal. Additionally, was involved in the support and enhancement of the following back-office automations of NCBC:

· Thomson & Reuters Data Automation

· Fund Management & Descriptive Portfolio Management Automation

· ALMANARAH

· Standard & Poor Data Automation

· Management Information System maintenance

Technologies: SharePoint 2007, MS SQL 2005, .NET 3.5, Windows Batch Scripting.

Gen-V Reservoir Simulator Project Details: The project is called Gen-V (“V” the Roman numeral for five). This project is intended for building the fifth generation Reservoir Simulation software for Exxon Mobil. This project is very critical to the client because the client is planning to use the Gen V simulation software for many years going forward within the organization and this simulator is crucial for their future business.

The project has 3 crucial modules:

· The 2D Plots Module

· The Algorithms Module (also called the Scripts Module)

· Property Editor Module

I was involved in the development of the Property Editor Module and was responsible for profiling and optimizing all the 3 modules using .Net 4 Parallel Extensions. A brief description of these three modules and the challenges we faced in developing those modules are given below:

2D Plots Module: The 2D Plots module is used to visualize the reservoir data. The 2D plots module uses the algorithms from the algorithms module to perform certain calculations before visualizing the data in one of the many different forms. The 2D plots module is an independent control which can be plugged into any host application. Currently it is integrated into Property Calculator. In this module we have heavily used Infragistics controls.

Algorithms Module: We optimized 10 crucial algorithms used in the reservoir simulation. We used the .Net 4 Parallel extensions to make these optimizations. The challenges we faced on this track were to optimize the algorithms keeping the memory constraints in view. The algorithms required a large amount of memory so came up with a workaround to allocate more memory for the process we launched to run the algorithms. Also came up with alternate data parallel algorithms that can perform better than the previous implementations. Profiled all the algorithms and helped the team to optimize them. After the optimizations were done, profiled the algorithms once again and a detailed report was submitted to the client and got a good appreciation for my work.

Property Editor Module: A reservoir simulator has many components and each of its components have many properties under many categories. End users can manipulate these properties with the help of a property editor, and they can run different simulations on the new property data. The real challenge in this module was that there are two kinds of end users, C# developers who have good knowledge of C# and Domain experts who don’t have a clue about what C# is, they just know how to write mathematical expressions. We had to implement all the features like supporting Intellisense, supporting automatic method argument filling, automatic resolution of common errors like missing semicolons and all such things at the same time giving the C# end users all the freedom to pour in their programming expertise.

Technologies: Infragistics Net Advantage, WPF, .Net 4 Parallel Extensions, PLINQ, Concurrency Visualizer.

Digital Rock Physics using LBM Simulation: Have re-engineered and ported the Open LBM Flow (http://www.lbmflow.com/) Lattice Boltzmann solver on to CUDA platform thereby significantly improving the performance. OpenLBMFlow is an open-source Lattice Boltzmann solver that is capable of simulating 2D or 3D single-phase as well as multi-phase fluid flow.

Technologies: NVidia CUDA, ParaView data visualizer.

LM Parser Optimization: The LM Parser is a tool developed for parsing and retrieving information from legacy projects consisting of COBOL files, JCLs and BMS Maps. The information retrieved by the parser is then used in estimating the time and cost models for modernizing that legacy project. Re-Engineered the Legacy Modernization Parser tool to use .NET 4 Parallel Extensions and brought down the time for parsing from 5.5 Hours to around 38 Minutes. Was involved in requirement analysis, Re-Engineering strategy, development, and testing.

Technologies: .Net 4 Parallel Extensions, SQL Server

Monte Carlo API for GPU Clusters: Developed a native API using C++ and CUDA C that can help CUDA programmers to develop Monte Carlo applications targeting NVidia GPU Clusters more easily.

Technologies: NVidia CUDA, Parallel Patterns Library.

Parallel Extensions POC: For demonstrating the performance benefits of using .NET 4 Parallel Extensions, created this POC; a back-office bank application which executes the “Scheduled Funds Transfer” requests placed by the customers. The sequential application took 1.64 days (2358 Minutes) to execute 1 million transactions. After the optimization of the application with the .NET 4 Parallel Extensions it was able to execute 1 million transactions in 0.35 days (501 Minutes) which is 78.75% reduction in time.

Technologies: .Net 4 Parallel Extensions, SQL Server

HPC Cluster Development POC: For demonstrating the performance benefits of using Microsoft HPC Server 2008 R2 Cluster along with .NET 4 Parallel Extensions created this POC. Re-Engineered the “Parallel Extensions POC” to scale up using the MS HPC server 2008 R2 Cluster and got reasonable performance gains. Was responsible for the scaling up strategy, coding, and testing of this POC.

Technologies: Microsoft HPC Server 2008 Cluster Development, .NET 4 Parallel Extensions, SQL Server.

Volume Computation POC: This POC was developed for demonstrating the performance benefits of using NVidia GPUs for embarrassingly parallel Monte Carlo Simulations. The team was able to show ~130x performance gain on NVidia GPUs in comparison to the multi-core implementation. Was involved in the optimization of the application and designed the User interface for this POC.

Technologies: NVidia CUDA, .NET 4 Parallel Extensions.

Kirchhoff Migration POC: Migration is the basic image making process in reflection seismology. Kirchhoff Migration is a method of seismic migration that uses the integral form (Kirchhoff equation) of the wave equation. For demonstrating the performance benefits of using GPUs, developed an application that does Seismic Analysis using Kirchhoff’s Migration on the many-core NVidia GPUs as well as multi-core CPUs and showed a 10x performance gain on NVidia GPUs.

Technologies: NVidia CUDA, .NET 4 Parallel Extensions.