5 Proud Accomplishments in the Computer Science Journey
Embark on a journey through the pinnacle achievements in computer science as this article unfolds key milestones and expert insights. Discover the monumental task of migrating over 80 applications to the cloud, the inception of a cutting-edge B2B prospecting platform, and the craft of developing an app from the ground up. Gain exclusive knowledge from industry leaders on leveraging AI expertise and creating a custom e-commerce site that serves millions.
- Leading Cloud Migration of 80+ Applications
- Building SpeakerDrive: A B2B Prospecting Platform
- Developing First App from Scratch
- Applying AI Expertise Across Multiple Companies
- Creating Custom E-commerce Site for Millions
Leading Cloud Migration of 80+ Applications
One of the most rewarding accomplishments in my computer science journey has been leading the successful migration of over 80 applications to the cloud at Optum. This experience was transformative, not just for my technical skill set but also for my leadership abilities, as it required harmonizing technological innovation with strategic vision to meet our business objectives.
The primary challenge in this endeavor was overcoming the complexity of migrating diverse applications, each with its unique dependencies and legacy systems, to a modern cloud infrastructure that required seamless integration and operational efficiency. This task was further complicated by the need to ensure minimal disruption to ongoing operations while achieving high availability and auto-scaling capabilities in the cloud.
To manage these challenges, I spearheaded a meticulous planning process, collaborating extensively with cross-functional teams to ensure a deep understanding of both technical requirements and business goals. We implemented a phased migration strategy, allowing us to methodically transition applications while continually optimizing performance through techniques like sharding, indexing, and caching, particularly using MongoDB Atlas after transitioning from Cassandra.
A critical component of our success was the development of robust CI/CD pipelines with Jenkins, which facilitated reliable and continuous deployment, significantly reducing the risk of errors and increasing our responsiveness to changing requirements. Additionally, leveraging technologies like Docker and Kubernetes, we streamlined workflows and improved operational efficiency, paving the way for real-time streaming solutions with Kafka and Micronaut microservices.
This achievement was not without its hurdles, particularly in terms of coordinating across large teams and ensuring alignment with overall strategic goals. However, these challenges were instrumental in honing my leadership skills and underscored the importance of a cohesive team effort in pioneering technological change.
Reflecting on this accomplishment fills me with pride, not just because of the technical milestones we achieved, but because of the tangible impact it had in enhancing scalability, efficiency, and resilience in our systems. It solidifies my belief in the power of integrating advanced cloud technologies with strategic foresight, a principle I continue to champion in my role as a Principal Engineer.
Building SpeakerDrive: A B2B Prospecting Platform
One of the most meaningful accomplishments in my computer science journey has been building SpeakerDrive--a specialized B2B prospecting platform for experts and speakers. It's not just a business tool--it's a CS-driven system built from the ground up to solve a messy, high-friction data problem using clean architecture and scalable infrastructure.
At its core, SpeakerDrive is a mix of information retrieval, data engineering, and applied machine learning. We built custom web scrapers in Python using Scrapy to crawl event sites, training portals, and niche directories. Then came the real computer science challenge: entity resolution and classification. We had to match unstructured speaker opportunities with the correct organizations, extract verified decision-maker contacts, and filter out noise. I used natural language processing (spaCy + custom rules-based parsing) to extract and classify roles, event types, and verticals.
We used MongoDB for flexible schema storage (since events and contact structures are wildly inconsistent) and deployed using Dockerized microservices on AWS ECS to scale ingestion pipelines. On top of that, we built a ranking algorithm--essentially a lightweight recommendation engine--that scores opportunities based on speaker compatibility using logistic regression trained on prior conversions and match data.
The biggest tech hurdle wasn't getting data--it was building trustworthy data. Speakers don't want 5,000 cold leads--they want 5 that actually book. So we implemented a scoring algorithm that factors in past speaker listings, event frequency, decision-maker availability, and industry relevance. That prioritization engine was a turning point. It took a lot of trial and error (and swearing at edge cases), but it paid off in retention.
Good CS isn't about shiny algorithms--it's about engineering reliable systems that solve messy, real-world problems. And that's exactly what we did. SpeakerDrive isn't just working--it's working well because of the deliberate CS decisions baked into every layer.
Developing First App from Scratch
One achievement that sparkles in my memory from my computer science journey is building my first app. It was both a thrilling and draining experience, especially because I was tackling it independently with no prior experience in app development. The biggest challenge was learning the coding language from scratch — every bug and error seemed like a huge mountain at that moment.
I remember spending nights poring over forums, watching endless tutorials, and experimenting with different solutions. When I finally launched the app and saw people actually downloading and using it, the feeling was indescribable. I learned the importance of persistence and patience, which are key components in the world of programming. Seeing a project come alive from just an idea to something people interact with was truly rewarding. This accomplishment has fueled my passion for technology and innovation, motivating me to take on even more challenging projects in the future.
Applying AI Expertise Across Multiple Companies
During my tenure at Capital One, I gained valuable experience in deploying large-scale machine learning models, improving efficiency, and optimizing pipeline deployment, which I applied at Freshworks. Key lessons included effective MLOps practices, reducing errors in production, and speeding up model deployment, which were instrumental in developing robust AI solutions at Freshworks.
At SparkCognition, leading data-driven projects in industrial IoT and prioritizing resources for product development taught me how to manage AI projects effectively and collaborate across teams. This experience helped in coordinating large-scale AI initiatives and ensuring efficient resource allocation at Freshworks.
Overall, these roles honed my skills in deploying impactful AI solutions, managing projects, and driving innovation, which I carried forward to enhance Freshworks' AI capabilities.
Creating Custom E-commerce Site for Millions
I studied Computer Studies at A Level in the late '80s and also got a BTech in Computer Science. I have always been involved with computers for work. In the 2000s, I started and ran quite an impressive and successful e-commerce shop in the UK. It was at this time I had an idea for an affiliate website, and for the last 17 years, I have tried several times to get this working using off-the-shelf software such as WordPress, Shopify, and WooCommerce. But the problem was that the site would be huge, up to 20 million products, so these off-the-shelf products simply can't cope with this volume of data. Most fall over around the 50,000 product mark. The main problem was importing the products into the DB, checking if it's new or to update existing data.
So in 2023, I decided that a custom site was the only route, and using the knowledge of the previous failures, I designed a system that didn't need the power of a server farm, so it was cost-effective as a startup. I put the project out to tender, agreed with the chosen developer, and 11 months later it was launched.
The biggest accomplishment was when the developers showed me the back-end import process that could churn through millions of products in minutes using my import design. I was sitting at my kitchen table when they demonstrated the system - I cried and my hands were shaking! It's early days since launch, so it's fingers crossed and hope for success in the future. If it makes money, great! If it doesn't, it will of course be disappointing, but I most certainly have a huge feeling of success in getting my idea actually working after so many years.
