“Managing engineers here means you set the context, then get out of the way and watch them run.”
Pradeep Anand Rathnavel, Machine Learning Engineer
- 01
What's the biggest shift in how you work over the last year?
I moved from building models to orchestrating them. The difference is significant. Training a model is a research problem with a six-week feedback loop. Orchestrating models in production is an engineering problem with a six-minute feedback loop. I run experiments on live recommendation traffic and see results before lunch. The iteration speed changes how you think.
- 02
Walk me through a problem you solved using AI that you couldn't have tackled the same way before.
We had a cold-start problem in merchant recommendations — new merchants with no signal. I used an LLM to synthesize pseudo-behavioral features from their onboarding data and category metadata. Zero historical clicks needed. The lift on new-merchant conversion was 14% in the first A/B test. That's not something you can brute-force with collaborative filtering.
- 03
What do you wish you'd known before joining?
How fast the stack changes. I came from a place where the ML infrastructure was three years old and stable. Here, the toolchain I use today didn't exist in the form it does when I joined. You have to be comfortable with the ground moving. If you treat that as energy rather than friction, you're well suited for this.
See open seats.
Building across engineering, product, data, sales, ops, finance, and people. Every role is an Operator role.
We get people offline through quality local experiences at great value. That's still the mission. Everything above is what it takes to deliver it in 2026.