{ "name": { "first": "Nicholas", "middle": null, "last": "Kelly" }, "about": "Senior Software Engineer with 10 years of experience in backend development, specializing in Python and Django. Skilled in building scalable web applications, optimizing performance, and implementing continuous delivery for efficient deployments. Passionate about creating innovative, reliable, and user-focused software solutions.", "position": "Senior Software Engineer", "experience": [ { "company": "Activated Insights", "position": "Senior Software Engineer", "timeperiod": "since March 2023", "description": [ "Built SyncStream, a scalable Python-based integrations framework that automated onboarding of new data sources, reducing implementation time from weeks to days and accelerating delivery of product features powered by ingest pipelines", "Led a 3-month rollout of a cloud data platform that unified backend data workflows across 8 products, improving reliability and enabling large-scale processing and analytics", "Improved performance of critical processing pipelines, reducing execution time by 73% and consistently completing workloads within a 15-minute SLA", "Strengthened engineering practices by integrating Codecov into CI/CD pipelines, improving visibility into test coverage and reinforcing automated testing standards", "Maintained and scaled distributed backend services using Python, AWS, Docker, and Terraform, focusing on reliability, observability, and extensibility of core backend systems", "Documented engineering workflows and system architecture to support cross-team collaboration and faster onboarding of engineers" ], "website": "https://activatedinsights.com" }, { "company": "Carta", "position": "Software Engineer II", "timeperiod": "January 2022 - January 2023", "description": [ "Designed and implemented a Python gRPC microservice for Faster Payments by integrating with the Plaid API, enabling faster payouts for 60% of customer transactions", "Reduced payment failures by introducing real-time account balance verification, lowering settlement risk and achieving 99.8% accuracy", "Built backend integrations with the Brex API to automate fund administration workflows, improving role assignment accuracy to 93%", "Implemented SonarQube branch analysis across team repositories to enforce code quality standards and maintain 76% test coverage", "Developed Python service layers to support Capital Call Line workflows, enabling reliable message transport between backend systems" ], "website": "https://carta.com" }, { "company": "Kellton Technology", "position": "Software Engineer", "timeperiod": "August 2016 - January 2022", "description": [ "Introduced an end-to-end testing framework (Cypress) across multiple applications, reducing regression bugs by 90%", "Modernized legacy Django applications using Docker, Ansible, CircleCI, and React, improving deployment reliability and reducing build times by 50%", "Built asynchronous Python services using Celery to ingest and process MLS data from multiple sources", "Developed a high-volume address resolution service using Python and Elasticsearch, reducing batch processing time by 60% and improving query performance at scale", "Improved user-facing performance by replacing Django template views with a React-based SPA, reducing page load times by 80%", "Converted a legacy scraping tool into a production Django application, enabling analysts to collect and model real estate market data" ], "website": "https://kelltontech.com" } ], "education": [ { "degree": "Bachelor of Science, Computer Engineering", "timeperiod": "January 2011 - December 2015", "school": "San Diego State University" } ], "skills": [ { "Languages/Frameworks": [ "Django / REST", "FastAPI", "NodeJS", "Python", "React", "TypeScript", "Spark" ] }, { "Infrastructure & Tooling": [ "Airbyte", "Codecov", "GitHub Actions", "Kubernetes", "LocalStack", "Grafana" ] }, { "Cloud & DevOps": [ "ArgoCD", "AWS (Lambda, SecretsManager, Athena, CloudWatch, EventBridge, Glue, S3, SQS, EventBridge)", "Docker", "Terraform" ] }, { "Databases": [ "Elasticsearch", "Iceberg", "MySQL", "PostgreSQL", "Redis" ] } ], "projects": [ { "name": "SyncStream", "stack": [ "Lambda", "Django", "Python", "Kubernetes", "Sentry", "Docker", "Terraform", "S3/SQS" ], "timeperiod": "March 2026", "company": "Activated Insights", "description": "I helped design and build SyncStream, our next-generation integrations platform, to replace a legacy engine built on an expensive no-code integration tool. While the original system enabled early integrations, it became difficult to extend, debug, and operate, leading to slow development cycles and limited observability. We rebuilt the platform using modern Python tooling and AWS services, implementing event-driven pipelines and durable storage in S3 to reliably ingest and process data across multiple products. By moving to a code-driven framework, we improved debuggability, reduced operational costs, and significantly accelerated the development of new integrations.", "website": "https://syncstream.activatedinsights.com/docs" }, { "name": "Data Lakehouse", "stack": [ "AWS Glue", "Airbyte", "Python", "Iceberg", "Kubernetes", "Sentry", "Spark", "Docker", "Terraform" ], "timeperiod": "March 2025", "company": "Activated Insights", "description": "One of the most impactful 0-to-1 projects I led was building a comprehensive data lakehouse from scratch, despite having no prior data engineering experience. I was tasked with creating a centralized data platform to unify analytics across eight different SaaS products. I began by conducting extensive research into lakehouse architecture, enrolling in a Udemy course on building data lakes with AWS to rapidly acquire the necessary expertise. I transformed this knowledge into actionable insights by drafting a comprehensive design document, which received leadership approval. Subsequently, I developed a proof of concept utilizing Docker, FastAPI, Airbyte, and AWS services. I collaborated with our principal architect to deploy cloud infrastructure using our already established IaC framework. To scale the initiative effectively, I spearheaded the development of a Python framework that enabled our team to onboard new data sources in days rather than weeks. The final outcome—we successfully delivered the complete lakehouse solution months ahead of the original schedule.", "website": null }, { "name": "Survey Sample Pool", "stack": [ "Django", "Django Rest Framework", "Postgres", "AWS (EventBridge, SQS)", "Terraform", "Docker", "ArgoCD", "GitHub Actions", "Kubernetes" ], "timeperiod": "August 2024", "company": "Activated Insights", "description": "Comprehensive participant pooling platform designed for survey participants, patients, and caregivers. Processes incoming webhooks from external vendors when patient data is updated or modified. The Django microservice architecture processes webhook events and publishes notifications to downstream consumers. This project marked my first experience developing a shared Python module utilized across multiple teams, implementing event-driven service architecture for scalable data processing and real-time communication.", "website": null }, { "name": "Integrations Platform 1", "stack": [ "TypeScript/Node", "TurboRepo", "JSONForms", "GitHub Actions", "ChangeSets", "Prismatic" ], "timeperiod": "January 2023", "company": "Activated Insights", "description": "Framework built on top of Prismatic iPaaS solution designed to significantly reduce implementation time for new customers. This was my first major project at the new company and marked my initial experience building with TypeScript using a monorepo architecture. The core engine was strategically separated from the client code to maintain vendor independence and flexibility. Established a complete CI/CD pipeline for both core library and client applications. Client applications automatically deploy new integrations upon code changes, while the core library generates new releases when modifications are made. Implemented strict 15-minute processing time limits, requiring optimization of all asynchronous operations to ensure fast data processing and real-time responsiveness.", "website": null }, { "name": "Scop", "stack": [ "Django", "Django Rest Framework", "Postgres", "Docker", "CodeFresh", "Go", "Kubernetes", "React" ], "timeperiod": "February 2021", "company": "Kellton Technology", "description": "Platform enabling real estate analysts to gather, analyze, and model data for properties on the market or under consideration for purchase. Developed this application over 3 years through 7 iterative versions. The project began as a simple web scraper deployed on an EC2 instance using crontab and BeautifulSoup4. The company recognized its value and requested a full web application. As the platform grew, we expanded to include additional microservices and React-based single-page applications. Significantly reduced analyst workload by automating the process of determining comparable properties in specific areas, including rental value assessments for potential property flips.", "website": "https://www.imagine-homes.com/rent" }, { "name": "Terralytics", "stack": [ "Django", "Postgres", "ExtJS", "Docker", "Ansible", "Vagrant" ], "timeperiod": "February 2018", "company": "Kellton Technology", "description": "Comprehensive platform for stakeholders to manage multi-scale information required for sustainable land-use management activities. This project significantly challenged my ability to adapt to a new frontend framework. Prior to this project, I had experience with Angular, React, and jQuery. The framework, ExtJS, was considerably outdated and lacked comprehensive documentation. However, I successfully learned the framework quickly and implemented new features according to customer requirements. Beyond the framework challenges, this was my first experience working with raster files, image overlays, and vector data processing. I enhanced the development workflow by migrating from Vagrant to Docker and implementing automated deployment processes using Ansible.", "website": "https://www.terraglobalcapital.com/terralytics-overview" }, { "name": "Dossier", "stack": [ "Django", "Django Rest Framework", "Postgres", "React", "GCP", "NLP", "Docker and Swarm", "CodeFresh" ], "timeperiod": "February 2018", "company": "Kellton Technology", "description": "Document management system for dossier storage and retrieval, serving approximately 10 leadership users. This project marked my first experience implementing SAML authentication within the Django authentication framework. As the lead developer, I collaborated with colleagues in Poland to deliver the solution. A significant learning outcome from this project was developing a comprehensive CI/CD pipeline using Docker, Docker Swarm, and CodeFresh (predecessor to GitHub Actions). Additionally, I integrated Google's Natural Language Processing service to extract information from documents and implement sentiment analysis features for the frontend users.", "website": null }, { "name": "Combinati", "stack": [ "Django", "Django Rest Framework", "SQLite", "React", "Electron", "Docker" ], "timeperiod": "February 2017", "company": "Kellton Technology", "description": "End-to-end digital PCR solution used by scientists to track disease-driven biomarkers and learn crucial information about the healthcare of individuals. This was an interesting project in that there was no public web component. The application required to be run on desktop (Windows) and needed to interface with custom hardware. Thankfully, the customer had a dedicated developer who was responsible for the software that bridged our application with their hardware. While I led the team on the project, I was responsible for the backend REST API and my colleagues in Poland were responsible for the frontend. This was the first project where I was able to work with a frontend developer and really define requirements on how RESTful endpoints should be structured.", "website": "https://ir.thermofisher.com/investors/news-events/news/news-details/2021/Thermo-Fisher-Scientific-Adds-Digital-PCR-to-Genetic-Analysis-Capabilities/default.aspx" }, { "name": "CloudVelox", "stack": [ "Django", "jQuery", "Vagrant", "Ansible" ], "timeperiod": "February 2016", "company": "Kellton Technology", "description": "CloudVelox was a SaaS solution for migrating on-premises infrastructure to the cloud (prior to being acquired by VMware). While I was not the lead developer on this project, I gained valuable exposure to cross-company collaboration and team dynamics.", "website": "https://www.thirdpointventures.com/companies/cloudvelox/" }, { "name": "Omus", "stack": [ "Angular", "C++", "PIC32", "GPIO" ], "timeperiod": "December 2015", "company": "San Diego State University", "description": "This was my first production web application deployed for a real audience with specific requirements. I was energized after being mentored by my supervisor who introduced me to Django and modern web technologies. My senior design project required our team to build an automated sumo bot from scratch and deploy a comprehensive website to document our development journey. The interdisciplinary team consisted of 2 computer engineering and 4 electrical engineering students. I served as the lead engineer for both embedded hardware systems and the web application, while my teammates were responsible for fabricating all the mechanical components. The website was built with Angular (deployed via FileZilla), while the autonomous bot was programmed in C++ with custom algorithms for movement, sensing, and combat strategies. The complete project code is available on my GitHub repository.", "website": "https://www.youtube.com/watch?v=eY6WTlivm7I" } ], "contact": { "email": "nialexanderke@proton.me", "phone": null, "street": null, "city": "San Francisco", "website": "nickkelly.info", "github": "nickatnight" } }