[Your Name] · [Email] · [Phone] · [City, ST]
April 21, 2026
Dear Hiring Manager,
I'm writing to apply for the Senior Software Engineer role on the Retail Buying Systems team. I've spent the last five years building inventory forecasting systems at Wayfair, and the work your team published on short-horizon demand signals at re:Invent is exactly the kind of problem I want to go deeper on — especially the trade-off between forecast accuracy and replenishment cost, which is where I've spent the most time.
At Wayfair I owned the forecasting service behind same-day fulfillment for 2.1M SKUs. The existing model was a batch ARIMA job that ran every 4 hours and missed demand spikes by hours, costing us an estimated $3.4M in out-of-stocks each quarter. I proposed and led a rebuild to a streaming gradient-boosted model on Kinesis + SageMaker that refreshes every 90 seconds, rolled it out SKU-by-SKU starting with the lowest-risk categories, and worked backwards from one question: what does the buyer on the other end need to trust this number? We cut out-of-stock rate by 31% and reduced excess inventory holding cost by $8.7M annualized, with zero customer-impacting incidents during the 14-week cutover.
Before Wayfair I was the second data engineer at a Series A logistics startup (Flexe), where I wrote the original warehouse allocation service and carried the pager for 18 months. That range — from being the only person debugging a Postgres lock at 2am to operating inside a 16,000-person org — is what I'd bring to Amazon. I've read the Leadership Principles closely, and the one I relate to most is Dive Deep; the forecasting rewrite worked because I spent the first three weeks sitting with the replenishment buyers in Boston, not in a Jira ticket.
I'd welcome the chance to talk about how I'd ramp up on the Retail Buying stack and what I think the highest-leverage problem to own in the first 90 days would be. I can share a write-up of the Wayfair forecasting rollout (STAR format, under a page) if it's useful for the loop.
Sincerely,
[Your Name]