← All guides

How I’d prep for Twitch’s Software Engineer I, Payments role in 30 days

Updated Nov 15, 2026

The job

Software Engineer I, Payments, Twitch

San Francisco, CA (in office) · $127,100 – $185,000 base, plus a sign-on bonus and Amazon RSUs · Posted Nov 13, 2026

View the posting →

What this team actually does

For a lot of streamers, Twitch is their job. The Payments team runs the system that gets them paid. Money comes in when viewers buy subs, gifted subs, Bits and Turbo, in lots of countries and with lots of payment methods. Money goes out once a month, when creators get their payouts. The posting says you’ll design and build Twitch’s commerce and payments platform and services, own your services end to end (“architecting, building and operating them”), work with product, design and data people on customer-facing launches, and write clear technical docs.

It’s an entry-level role: 1+ years as a software engineer.

AreaWhat to know
LanguageGo (the posting says Go, Ruby, Java or C++; the rest of Twitch’s Commerce org says Go)
DataDynamoDB: data modeling and schema design, conditional writes, transactions
Async workflowsSQS queues and Step Functions for multi-step jobs that retry
ComputeLambda (serverless) and ECS on AWS. Twitch is owned by Amazon
Service callsTwirp, Twitch’s own open-source RPC framework for Go
FundamentalsData structures and algorithms (it’s in the posting)
The stack, from the posting, Twitch’s other Commerce postings and Twitch’s engineering blog.

How I found the stack (do this for any job)

  1. Read the posting closely. It says “Go, Ruby, Java, C++” and “familiar with AWS”. That’s four languages and a whole cloud, so it doesn’t tell you much. One clue: the November version added “Experience writing Go in production” as a bonus. The August version didn’t have it.
  2. Check other postings from the same org. Payments sits in Twitch’s Commerce engineering org. An older Commerce posting, “Software Development Engineer, Money” (closed, still readable on Built In), and a live Commerce posting from the same week both name the same stack: Go, DynamoDB, Lambda, SQS, Step Functions and ECS.
  3. Read the company’s engineering blog. Twitch’s Breaking the Monolith series explains how Twitch moved off Ruby on Rails and into Go microservices. Their services talk to each other with Twirp, an RPC framework Twitch open-sourced.
  4. Look for the team on LinkedIn. Search Google for site:linkedin.com/in Twitch "Payments" engineer and read what people list for their current role. Don’t scrape LinkedIn; just read.

So: Go on AWS, with Lambda, DynamoDB, SQS and Step Functions. That’s where I’d spend the 30 days. AWS has over 200 services; these are the ones that matter for this team.

Understand the product first

Spend an hour using Twitch’s money features as a viewer and as a creator. It’s the fastest way to sound like you belong on the team.

  1. Subscribe to a streamer, then gift a sub to someone in chat. Notice the checkout, the receipt, and what happens when a sub renews.
  2. Buy the smallest pack of Bits and cheer with them. Bits are a currency inside Twitch, so someone has to track every balance.
  3. Open the creator dashboard (even on a brand-new channel) and find the payout settings: payout method, tax forms, and the revenue page.
  4. Read Twitch’s payout minimum post. Creators get paid monthly once they reach $50 (wire transfers still need $100), and anything under that rolls over to the next month.

Then ask yourself the questions this team answers every day: what happens if the card is declined on a renewal? If a viewer gets charged and the network drops before Twitch hears back? If a creator changes their payout method the day before payouts run?

System design questions

  • Design the payment flow for a Twitch sub. What happens if the payment provider times out? How do you make sure the viewer is never charged twice?
  • Design Bits: a currency viewers buy and spend in chat. How do you keep every balance correct when thousands of people cheer at once?
  • Design the monthly creator payout run for millions of creators. How do you make it safe to rerun if it crashes halfway?
  • A big streamer’s community gifts 10,000 subs in one minute. What breaks first, and how do you handle the spike?
  • Design a ledger. Why do payment systems record every movement of money instead of just updating a balance?
  • How would you model subscriptions in DynamoDB? Which access patterns drive your keys?

Concepts they’ll probe

  • Idempotency: idempotency keys, and why “retry until it works” is dangerous with money.
  • Transactions: ACID, isolation levels, lost updates, and race conditions on a balance.
  • DynamoDB: partition and sort keys, single-table design, conditional writes, TransactWriteItems, hot partitions.
  • Queues: at-least-once delivery, duplicate messages, dead-letter queues, SQS FIFO deduplication.
  • Workflows: retries with backoff, timeouts, and sagas (undo steps when a later step fails).
  • Ledgers: double-entry bookkeeping, and why you never update or delete a ledger row.
  • Webhooks: verifying signatures, handling events that arrive twice or out of order.
  • Go: goroutines and channels, context for timeouts, error handling, interfaces, writing tests.
  • Data structures and algorithms: hash maps, heaps, sorting, intervals, BFS/DFS.

Unlock the rest

  • The week-by-week 30-day plan, with daily practice on Devmaxx
  • The exact book chapters and docs for each week
  • The project: a fishing tournament manager in Go on Lambda, DynamoDB, SQS and Step Functions
  • The entry-level bar: what they expect from someone with 1–2 years
  • The resource list with every link

Get the full guide, free

Enter your email and the guide opens right here. You’ll also get the occasional Devmaxx email. Unsubscribe anytime.

Twitch Payments Software Engineer Interview Prep: A 30-Day Plan · Devmaxx