a coder's tale

by Robert Roskam

who's this guy?

contributor

organizer

I've worked places and done things.

  • Multi-national conglomerate
  • ~300k employees
  • ~20% of South Korea's GDP
  • 20+ year old

Tier 2 Social Media Support

  • Indirect Auto Lender
  • 200 employees; 80% in Greenville, SC
  • $2B in loans outstanding of the $1.3T
  • 5 year old VC backend M&A

Technical Analyst I

Projects

  • Migrate LOS from CMSI (OracleDb SP/XML) to DEFI (JS)
  • Automate testing for decisioning (Python)
  • Self-service password resets for various tools (C#)
  • Software Consulting Company
  • ~15 employees; 50% in Greenville, SC
  • ~100 customers across the US
  • 25+ yrs old

Developer

Sample of Projects

(50+ projects grand total)

  • Conference Room Reservation System
  • Benefit Tracking System
  • Call Center System

Lead Engineer

  • Wellness Provider
  • ~100 employees; fully distributed in US
  • $40M in VC funding
  • 5 year old start up

Backend Engineer

🏝 Work remotely

⛺️ Unlimited Paid Time-off

📈  Company Stock options

Team Lead

  • Email Security
  • ~40 employees; fully distributed US, CA, NL, BG
  • No VC Funding
  • 5 year old start up

Staff Engineer

Lessons/Thoughts

Good Code ↛ Success

  • You can make your org fail with bad code, but you can't succeed on your own.
  • The code you're not proud of will stay around forever, because it works, so no one touches it.

Soft Skills are Hard

  • Companies that value written communication above oral communication are rare and nearly always winners.
  • Low trust environments require lots of communication. If people need to talk a lot to get "alignment", it's because everyone is new or everyone distrusts each other.

Your Specific Skill of X is not Valuable across Companies

  • You can find jobs as just a C# person, but this is also true of COBOL
  • Learn to have your skills translate across tech stacks

Old Tech ↛ Bad Tech

1993

1990

Success feels Boring

$40M in funding

unprofitable

5 years old

Customers in US

1 DC Center

40 engineers

No funding

profitable

5 years old

Customers worldwide

4 DC Regions

10 engineers

Relationships Last Across Jobs

More Pithy One Liners

  • Broke gets fixed; crappy is forever
  • Code ages like fish; data ages like wine
  • Today's experimental greenfield is tomorrow's legacy deathtrap
  • Every company generally believes they're following best practices
  • Quantity of outstanding bugs is a function of usage throughput

contact info

robertroskam.com

me@robertroskam.com

 

@raiderrobert 

twitter | github | medium

GVL Tech Talk September 2019

By Robert Roskam

GVL Tech Talk September 2019

  • 165