Rich Seviora

Responsible for the overall stability and direction of the Rendering platform, and the next generation ordering platform.

Managed multiple cross-team, cross-system multi-quarter projects including:

• Order Failures - led the Lob staff engineer group in designing the ability to automatically fail orders across multiple systems and provide actionable information to the customer. This lead to a ~99% reduction in order defects and helped retain a critical enterprise customer.
• Shadow Pipeline - key contributor to the design and implementation of a shadow ordering pipeline, allowing Lob to identify potential rendering fidelity issues across a wide variety of input creatives in advance of releasing code changes to production. This generated a >95% reduction in rendering fidelity incident frequency and $200K in annual savings.
• Order Deduplication - designed the ability to identify orders with duplicate creatives and coordinate system and process changes so that print partners can only download one creative. This reduced S3 order costs by approximately 40% and virtually eliminated partner download time for these campaigns.

Iteratively improved the performance and cost effectiveness of a large distributed PDF rendering pipeline across a combination of AWS Lambdas and containerized services (primarily NodeJS, with some Go), handling highly variable input volumes and sources:

• Reduced per unit rendering compute costs by 30-40% while maintaining a prompt rendering experience (99% rendered in under 60 seconds).
• Reduced S3 costs by $250K/year.

Key contributor to the upcoming GA launch of the next generation ordering platform with a seamless API experience for existing customers.

Designed the core structure and patterns of the next generation PDF production engine, enabling other engineers to add new features sustainably and quickly.

Provide coaching and guidance to client engineers, participated in training and development for Test Double consultants.

Lead engineer on the Features and Capability team developing the Olive Helps AI copilot software.

• Ideated and designed all key desktop sensing capabilities to satisfy prospective developer needs.
• Created a custom React renderer, allowing Loop developers to create complex user interactions using the full React API within the Olive Helps custom JS runtime. This reduced code required to manage complex interactions by upwards of 50%, enabling Loop developers to produce higher value experiences.
• Contributed to the development of the Olive Helps design system and UI components made available in Olive Helps Loops.
• Created and maintained first generation SDKs for multiple frameworks/languages (Node, Go, .Net).
• Designed the core SDK architecture when Olive AI adopted a secure JS runtime to execute their Loops (customer/vendor provided code).
• Learned Go on the job to better support the team; becoming proficient within a matter of weeks.
• Collaborated closely with product management on the design and prioritization of new Olive Helps features and capabilities.
• Trained and onboarded dozens of engineers to the Olive Helps platform.
• Created and delivered multiple training sessions on Typescript and Olive Helps to large audiences.