Topline Engineering Roadmap

Principles:

1. Privacy and Security. Users expect it, regulations require it.

2. Integrated. Users expect the ability to connect to other services.

3. Responsiveness. The system needs to be nimble and responsive. Also at scale (below).

4. Scalability. With growth, potentially explosive, the system must remain available and responsive.

5. Agile. Product will develop, often responding to users’ feedback. The development cycle needs to be very short.

6. Maintainability. Ensuring that the codebase is well documented and readable, with rapid understanding and onboarding for new hires.

CULTURE

Topline’s engineering culture will be to foster the following principles:

1. Optimize for programmer happiness

   1. Technology choices should empower engineers to do more with less.

2. Convention over configuration

   1. Developers should not be responsible for re-engineering best-practices 

3. Simplification 

   1. Getting our work to the user as early as possible

4. Empowerment

   1. Giving engineers the tools the need and the freedom to complete their work independently.

5. Rigor

   1. Expecting engineers to follow all paths to solve problems and apply reason to achieve successful outcomes.

6. Data Driven

   1. Ensure that user feedback is captured via voting, regular Q&A with product leaders, and usage data as a major component of product roadmap

   2. Ensure each goal has associated metrics or key performance indicators (KPIs) to monitor progress and success. How will you measure privacy, integration, responsiveness, scalability, agility, and maintainability objectively?

PROPOSED TECHNOLOGIES RATIONALE

The intended users of Topline are likely to be accessing the tool using their desktop or mobile phone. Accordingly, the infrastructure should be simplified to allow for deployment to iOS, Android and Web with minimal code maintenance. 

Business Logic

While Topline exists to present data in a CRUD ( create, read, update, destroy ) fashion, a decent amount of background processing is expected to engage users and to automate connections between users.

Peripheral Systems

Employ components, integrations and external services for anything that’s not core business:

• CyberSecurity (external) - CloudFlare

• Performance - Cloudflare

• User support - Gleap

• Internal Bug Reporting - Jam

• User behavior tracking - Google Analytics and PostHog

• Email gateway - Mailgun

• Payment gateway - Stripe

• Asset Storage - AWS S3 or similar

• Error Tracking - Sentry

• Performance Tracing - Scout

• Data Storage - Postgres

• Caching - Redis

SYSTEM ARCHITECTURE DESIGN

Topline will utilize a monolith web application architecture to support its web services to users. The monolith will combine data storage, caching, background web processing and notifications, with integration to 3rd party providers for ancillary services. It will use client web wrappers for the iOS and Android native clients. Topline will leverage Jumpstart Pro as the SAAS wrapper, to avoid writing boilerplate code. 

Technologies

1. Backend

   1. Rails, Ruby

2. FrontEnd

   1. Slim, Hotwire, Stimulus, Turbo 8 (Server Rendered Views)

3. Background workers

   1. Sidekiq, Redis

4. Email

   1. Mailgun

5. Application Monitoring

   1. Scout

6. Error Tracking

   1. Sentry

7. Static Asset Hosting

   1. Amazon S3

8. Application Hosting

   1. Cloud66 / AWS -> Heroku

9. SEO

   1. FriendlyID slugs (e.g. withTopline.com/tyler-horan)

   2. AI driven meta tags with semantic entity parting ( Google ML ).

   3. Automated Sitemaps with automated pings delivered to Google, Bing & Baidu.

      1. Structured Snippet with Topline Score

10. Internationalization

    1. i18n based on browser locale (en, es, fr, zh)

    2. Translations from Google Translate. Future versions use human translations.

11. Testing

    1. MiniTest, Capybara, UserTesting

12. Penetration Testing

    1. Pentest (https://github.com/hakatashi/pentest) 

13. Advertisement

    1. Google, Facebook and Baidu tracking pixels inside Cloudflare Zaraz

14. Quality Assurance

    1. Use of Review Apps to stage important feature development. Documentation of features in Notion and use of manual QA testing on non-standard work hours contractors.

15. Support

    1. Gleap Support Built In

    2. Email Mailto from application creates tickets on notion project board.

    3. Admin and Support Team can Impersonate a user to debug issues.

16. Feature Flagging

    1. Exponential Rollout to percentages of users on major releases.

       1. Day 1: 1%

       2. Day 2: 4%

       3. Day 3: 16%

       4. Day 4: 64%

       5. Day 5: 100% 
          

Technology Rationale

1. Choosing boring technology

2. Choosing technology where the output per engineer is highest. 

Alternative Evaluated Choices

1. Backend

   1. Javascript/NodeJS

      1. Popular, large talent pool. Lower output per engineer.

   2. Laravel/PhP

      1. No strong support for multiple platforms 

      2. Pay Lock in for support on modules.

2. Front End

   1. React

      1. Large library of UI components, but separation of state requires a lot of additional developer overhead in separating front end from back end.

   2. Vue

      1. Similar structure as React, but lacking the backing of larger enterprises like Facebook.

   3. Angular

      1. Large library of UI components, but separation of state in bi-directional flow complicates the web application.

   4. Ember

      1. Legacy framework with a much smaller pool of application developers.

3. Background Processing

   1. Resque

4. Email

   1. Similar offerings are available with other providers, but require account provisioning outside of the PAAS ecosystem. These include:

   2. Amazon Simple Email Service

   3. Mailgun

   4. Postmark

   5. SendinBlue

5. SMS

   1. Similar offerings are available from a number of providers. Twilio has better documentation and the pricing difference at low volume is marginal.

   2. Plivo

   3. Vonage

6. Application Monitoring

   1. Skylight

   2. New Relic

   3. Datadog

7. User Tracking

   1. Using third party analytics services makes it more difficult to be GDPR and California compliant. But a number of third party tools are available.

      1. Google Analytics

      2. Heap 

8. Error Tracking

   1. All the providers below offer similar tooling and monitoring of the application.

      1. Sentry

      2. Rollbar

      3. Airbrake

      4. Raygun

9. Asset Hosting

   1. AWS S3

      1. Raw usage of cloud providers requires maintaining additional accounts, a security vulnerability.

   2. Google Cloud Storage

      1. Similar performance to Amazon. Raw usage of cloud providers requires maintaining additional accounts, a security vulnerability.

   3. Azure Storage

      1. Similar performance to Amazon. Raw usage of cloud providers requires maintaining additional accounts, a security vulnerability.

10. Application Hosting 

    1. Bare metal hosting of the application using IAAS (Infrastructure as a service), should be avoided in favor of PAAS. 

       1. DevOps professionals are required to maintain infrastructure otherwise. 

    2. Google Cloud

       1. Compute Engine

       2. App Engine

    3. Amazon Web Services

       1. Engine Yard

       2. Beanstalk

       3. EC2

    4. Azure

       1. App Service

       2. Virtual Machine
          

Engineering Org Structure Evolution

1. 2024 

   1. Flat structure with majority of development work completed by CTO early in the year. Following completion of initial V1 build, development work should be parallelized by allowing additional engineering talent to take ownership of a particular feature dimension.

2. 2025 

   1. Addition of VP of Eng or Engineer Manager to organize multiple squads (teams) into agile sprints. Each squad should be oriented around development of a particular feature dimension. Squads can be spun down, re-organized and spun up depending on company ORK and goals for the year. 

3. 2025- 2030

   1. Expansion of engineer’s per squad, with the addition of domain experts on a particular aspect of the technology stack (e.g. front end, back end, data science, etc.)
      

Hiring Roadmap

1. Q2 2024. NA, CTO leading application development to V1, augmenting development with contractors for easily compartmentable tasks when needed. (Figma to HTML, Automated MiniTests, etc.)

2. Q2-Q3 2024. 1-2 Full Stack Engineers. Each should take ownership of high impact product feature dimension

3. Q4 2024 - Q1 2023. 1-5 Full Stack Engineers. Evaluation of Product Roadmap to determine level of application support needed to meet business goals.
   

Hiring Sources

1. CTO personal networks

2. Conferences - RailsConf, RubyConf

3. Job Boards - Topline, Turing, TopTal, Hired, HackerNews, AngelList
   

Hiring Screening Process

1. Interest conversation w/ CTO

2. Technical project

   1. ½ take home project

      1. Testing ability to create an app from scratch that:

         1. Pulls data from an endpoint

         2. Stores data from a form

         3. Customizes the view of data presented to the user.

         4. Highlights front end, back end and network system design.

   2. ½ project pairing with CTO or other engineer

      1. Continuation of the take home project to full completion.

3. Systems Design conversation with engineer 

4. CEO and non-eng culture fit conversation