Projects

Tech E&O Dynamic Applications Project

• Spearheaded a strategic initiative to transition insurance application questions from static to dynamic, starting with Tech E&O. This project was critical to enabling customized insurance products, directly impacting business agility and the ability to enter new markets faster.
• Drove the architectural vision by creating the initial design, facilitating cross-team architectural discussions, and presenting multiple options with clear trade-offs to leadership. My ability to build consensus and "disagree and commit" ensured the project moved forward with full business alignment, even when the final direction differed from my initial proposal.
• Led the technical execution, integrating our document services with the exec-risks-backend under tight deadlines. This involved navigating a complex and unfamiliar codebase and personally implementing a significant portion of the core logic, ensuring we delivered a high-quality solution on time.
• Championed and executed a critical cross-team refactor to centralize Protobuf definitions. I identified inconsistencies in the exec-risks-backend and led the effort to centralize protos across several services, reducing code duplication and simplifying future integrations. This has become the new best practice for services in our domain.

Policy Validator Service

• Leadership Role in Service Creation:
  • Successfully took ownership of the service, building upon its foundation and enhancing its capabilities, demonstrating initiative and problem-solving skills.
  • Demonstrated initiative and problem-solving skills by optimizing and extending the service beyond its initial scope.
• Expanding the Work and Reducing Noise:
  • Collaborated with key stakeholders and product teams to deliver meaningful metrics and safeguards, showcasing strong collaboration and communication skills.
  • Achieved higher coverage than promised within the quarter (60% promised, 70% delivered), showcasing dependability and over-delivering on objectives, illustrating dependability and an ability to exceed expectations.
• Identifying and Addressing System Issues:
  • Exposed several key issues, bugs, and gaps, leading to the creation of a multi-team project to address these problems, highlighting strong analytical and teamwork skills.
    • Worked through adversity to actually still deliver key goals
  • Gathered data and worked with a senior developer to better categorize and address these issues, illustrating strong analytical and teamwork skills.
    • (Naming is one of the two hardest things in SWE 😉)

Impact

• Increased system reliability and performance by identifying and resolving critical issues.
• Fostered cross-team collaboration and improved overall service quality.
• Enhanced project leadership and strategic problem-solving abilities.

Template Management Refactor

• Refactoring and Renaming:
  • Heavily refactored the template management logic, renaming the "dynamoregistry" to "template management" to provide a new framework for grouping related functionality, showcasing initiative and strategic thinking.
  • Split out logic for the local cache and DynamoDB (DDB) calls into separate providers by creating an abstract class, demonstrating strong design skills and modularity.
• Efficiency and Performance Improvements:
  • Improved DDB queries and scans by avoiding full table scans and introducing direct item queries, significantly enhancing the tool's performance and efficiency.
  • Developed a concept of a search_key for more refined scans, increasing the system's responsiveness and search functionality.
  • Extended search capabilities to include fuzzy search and filtering by various attributes such as market, line, and other configurable fields, paving the way for future performance enhancements.
• Real-Time Updates and Synchronization:
  • Ensured updates or creations in DDB are immediately saved and synced across instances, maintaining the health of the document management tool during deployments.
  • Improved system scalability by transitioning from a single instance to multiple instances, allowing as many instances as needed.
  • Enhanced system stability by eliminating the dependency on full table scans and improving cache reliability.

Impact

• Enhanced Performance:
  • Reduced the need for full table scans, significantly speeding up the doc management tool.
• Scalability:
  • Successfully transitioned from a single instance to multiple instances, providing the ability to scale out as needed.
• Improved System Health:
  • Ensured real-time synchronization across instances, reducing system downtime during deployments.
• Increased Modularity and Maintainability:
  • Made the logic more modular and easier to maintain by separating concerns into different providers.
• New Framework Implementation:
  • Provided a new paradigm for thinking about and grouping functionality, enhancing the clarity and strategic direction of the system.
• Advanced Search Capabilities:
  • Enabled fuzzy search and customizable filtering, increasing search efficiency and flexibility.

Replay Service

• Implementation and Testing:
  • Implemented a smoke/regression testing system with PE, improving the system's reliability and robustness through technical expertise.
  • Took significant ownership of infrastructure work, system debugging, and codebase contributions, demonstrating a deep technical expertise.
• Process Design and Training:
  • Designed an effective testing process and successfully trained other team members, showing strong teaching and leadership abilities.
  • Ensured minimal introduction of new bugs and improved testing comprehensiveness.
• Frontend Modernization:
  • Successfully migrated frontend from Jinja to React
  • Established new repository following company best practices
  • Implemented modern development tools and practices:
    • Integrated pnpm for significantly improved local development speed
    • Adopted tanstack-query for efficient data management
    • Developed custom hooks and optimized React Context usage
  • Created foundation for modernizing older frontend applications
  • Enhanced developer experience through improved tooling

Impact

• Increased system reliability with thorough smoke and regression testing.
• Elevated team productivity and code quality through structured training and robust testing frameworks.
• Demonstrated leadership and mentorship by guiding teammates through complex technical processes.

Document Management Tool Improvements

• User Experience Enhancements:
  • Improved the user interface by allowing users to customize which subsections they see in the conditions section, enhancing ease of use and user satisfaction.
  • Introduced customizable preferences for the conditions and main templates page. Users can now select and order columns based on their needs.
  • Expanded the variety of columns available, including market, line, and assignment status, improving the visibility of critical information.
• Collaboration and Feedback:
  • Actively collaborated with doc-support users to gather and understand their needs and preferences, ensuring the tool improvements were user-centric.
  • Responded promptly to user feedback and bugs during the rollout, addressing issues within minutes, and showcasing responsiveness and customer focus.
• Enhanced Document Version Control:
  • Implemented rollback endpoint for files, enabling version control in S3
  • Created endpoint for retrieving file versions and visual diff capabilities
  • Navigated S3 versioning constraints to implement a robust solution
  • Added critical document management features that improved service capabilities
  • Assisted doc support engineers with large PRs during company logo rollout

Impact

• Enhanced usability and user satisfaction by facilitating quicker and easier access to important information.
• Improved user efficiency and customization options, leading to a more productive workflow for doc-support users.
• Demonstrated strong collaboration, communication, and responsiveness skills, ensuring ongoing user engagement and improvement.

Docgen Service Latency Improvement

• Problem Identification and Collaboration:
  • Noticed significant latency issues in the main service, Docgen, which caused us to be way out of our SLO and significantly increased our burn rate.
  • Collaborated with a senior engineer who initially provided possible sources of the latency problem.
• Investigation and Analysis:
  • Took ownership of the issue while the senior engineer was away, investigating the suggested issues using logs and traces in Datadog.
  • Maintained ongoing communication with two Principal Engineers (PEs), sharing insights and collaborating on potential solutions.
• Solution Implementation:
  • Identified that reloading the local cache every minute was causing the latency.
  • Proposed and implemented a quick solution to update the cache only on changes, leveraging my recent work on the caching mechanism and the new system for capturing snapshots of our database.
  • Implemented a mechanism where the cache would check the snapshot ID and reload only if it had changed.

Impact

• Successfully restored our SLOs and significantly reduced latency, as evidenced by several graphs:
  • Reload time was reduced from several seconds to mostly 0s, with mere milliseconds on actual reloads.
  • The percentage of calls over 1s dropped from 30% to 10%, improving overall performance.
• Also fixed a longstanding memory issue, leading to a dramatic drop in memory usage.

TRIA Free Offering Project

• Leadership and Coordination:
  • Led a high-profile, customer-facing project to offer TRIA for free in a specific market for our internal carrier.
  • Coordinated with multiple teams and stakeholders, managing several setbacks, including requirements clarifications and carrier approvals.
• Handling Legacy Policies:
  • Managed the complexity of older policies and determined required behavior, leading to the identification and resolution of several bugs.
  • Collaborated closely with different teams to get engineering questions answered and disseminated information effectively.
• Cross-Team Collaboration:
  • Facilitated inter-team communication, ensuring that developers from various teams were aligned.
  • Specifically addressed a front-end issue where older policies were not accounted for, and worked with the front-end team to resolve it.
  • Quickly directed the resolution of bugs and ensured that customer issues were promptly handled.
• Process Improvement:
  • Introduced new processes for more effective project management, such as clear launch documents with specified dates and sign-offs, inspired by best practices from senior developers.

Impact

• Simplified User Experience:
  • Eliminated the need for policyholders to accept or reject TRIA with a signature, reducing hassle and streamlining the onboarding experience.
  • Updated internal tools to remove prompts for TRIA acceptance or rejection, speeding up the policy issuance workflow.
• Benefits for Brokers:
  • Enabled a smoother and simpler policy setup for brokers by removing the need for an additional onboarding discussion point.
  • Promoted consistent TRIA offerings across all US Cyber policies, enhancing the user experience.
• Improved Policy Documentation:
  • Updated policy documents (Quotation, Binder, Declarations) to no longer show TRIA charges, and revised marketing jackets to reflect this change.
• Operational Efficiency:
  • Simplified policy issuance workflows, allowing the servicing team to avoid TRIA-specific handling processes.

Collaboration & Mentorship

Collaboration

Designated Go-to Expert

• The Documents and doc-support teams regularly seek my expertise on Golang-related issues and questions.
• As the subject matter expert (SME) for infrastructure within the Documents team, I guide engineers of all levels—from P2 to senior—on setting up Terraform and Terragrunt files.
• Demonstrated strong mentorship and knowledge sharing:
  • Helped doc support engineers with large PRs during logo rollout
  • Provided rapid response to bug reports and feature requests
  • Assisted with giant PRs for the docx_formatter project
  • Provided mentoring support to doc-support engineers, fostering their growth and development
  • Shared React and modern frontend development best practices
  • Led by example in improving documentation and developer workflows

Mentorship and Co-Contributor Role in Design Docs

• Consulted PE and Senior Engineer on Multiple Projects
  • Mentored other developers, including guiding P3 engineers through a downturn and helping them relate to other team members.
  • Co-contributed to design docs and consulted with PE and a Senior Engineer on multiple projects.
• Push for more documentation and formalization of design decisions

Cross-Team Technical Leadership

• Tech E&O Project Leadership:
  • Created a comprehensive initial design document
  • Demonstrated adaptability by pivoting based on senior principal engineer feedback
  • Developed and presented multiple architectural options with clear trade-offs
  • Successfully integrated with exec-risks-backend despite architectural differences
  • Led proto file centralization initiative across multiple services
  • Completed complex refactoring work in external team repositories
  • Facilitated smooth handoff to exec-risks team under tight deadlines

Document Management Tool Collaboration

• Worked closely with doc-support users to understand their requirements, gather feedback, and implement improvements.
• Maintained rapid response times to user feedback and bug reports during the rollout phase.

General Recognition

• Consistently recognized by a Principal Engineer for exceptional responsiveness and quick fixes for any and all bugs:
  • "Thanks for always being so quick to make fixes and investigate things!" (Jim Kelly)
  • Multiple instances where the PE has been impressed by the speed and quality of the fixes delivered, underscoring my reliability and dedication to quality.
• Actively collaborated with teammates and stakeholders across projects, ensuring timely communication and effective problem-solving.

Docgen Service Latency Improvement

• Collaborated closely with a senior engineer and two Principal Engineers to diagnose and resolve significant latency issues in the Docgen service.
• Maintained clear and ongoing communication throughout the process, ensuring alignment and timely updates.

TRIA Coverage Offering Project

• Led cross-functional collaboration involving multiple teams to successfully launch the TRIA coverage offering.
• Managed and resolved issues related to older policies, coordinating with various teams to ensure engineering questions were answered and solutions were disseminated.
• Promoted better communication and project management practices, such as clear launch documents with specified dates and sign-offs.
• Addressed frontend issues regarding older policies and worked with the frontend team to resolve them, ensuring a seamless user experience.

Foundational Code

Refactoring and Documentation

• Refactored the main service code to improve readability and functionality, introduced better testing, and added comprehensive docstrings.
• Created foundational code for the Policy-Validator service, including structure and best practices.
• Improved several CI systems and developer workflows.

Infrastructure and Developer Experience

• Improved repository documentation and developer onboarding:
  • Enhanced READMEs with comprehensive documentation
  • Optimized makefile commands for better workflow efficiency
  • Improved docker compose configuration with network additions
  • Streamlined and accelerated CI processes

Design & Documentation

Design Docs

1. Wrote design docs for Tech E&O project—how the new integrations can work with tradeoffs for each

Documentation

1. Heavily improved READMEs in all repos owned

What you learned

Concepts and Books

Read several books:

Title	Key Takeaways
Software Architecture: The Hard Parts	Does a great job explaining complex issues, though the narrative tone falls short.
Fundamentals of Software Architecture: An Engineering Approach	More approachable than Kleppmann's DDIA, and very detailed and clear about key concepts.
The Staff Engineer's Path by Tanya Reilly	Provides helpful advice and practical ways to think and act like a staff engineer.
Good Strategy / Bad Strategy: The Difference and Why it Matters	A simple but mind-blowing idea that changed my perspective on a fundamental concept.

Concepts:

• Deep dive into TDD
• Deep dive into programming paradigms
• Dependency Injection (DI).

Libraries and Tools

• Gained expertise in modern frontend development tools:
  • Implemented pnpm for significantly improved local development speed
  • Mastered tanstack-query for efficient data management
  • Developed custom React hooks and context patterns
  • Learned advanced S3 versioning and management
• Deep dived into Bash
  • Moved beyond basic features of bash to more advanced features
  • Created several more user-friendly bash scripts
  • Practiced better error handling and responses in bash scripts
  • I learned a lot of cool tricks from https://www.youtube.com/@yousuckatprogramming

Outside of work

Blog posts

Programming Paradigms