Sana Nisha Joshi

ML Engineer at Mercari · ML infrastructure

Mercari

Auckland, New Zealand · NZST (UTC+12)

2+ years experience

Nanyang Technological University

English

Messaging unavailable

Overview

I've spent 2 years on ML infrastructure - first at McKinsey & Company, now at Mercari. Cut p99 latency on the ML platform services from 720ms to 85ms. Grew the ML platform team from 6 to 36 engineers. Most at home with PyTorch and CUDA. Collector of fountain pens.

Good reasons to reach out

Second opinions on Ray decisions

Please don't message about

Agency or outsourcing pitches
Offshore staffing offers

Slower in the last week of each quarter

Skills

RayPyTorchmodel servingfeature storesCUDA

Experience

ML Engineer
Mar 2026 - Present
Mercari · 7 mo · Full-time · Auckland, New Zealand
- Replaced 111 hand-built dashboards with one metrics layer people trust.
- Taught SQL to 189 non-analysts.
ML Engineer
Oct 2024 - Mar 2026
McKinsey & Company · 1 yr 5 mo · Full-time · Auckland, New Zealand
- Taught SQL to 118 non-analysts.

Projects

Write-up: recommendation systems in practice

2023

A long-form guide to recommendation systems for people starting out.

References

No references yet.

Education

Nanyang Technological University
BEng, Electronics and Communication
2019 - 2023

Details

Compensation expectations

What they are targeting for their next role.

Visible to connections

Direct contact details

Direct email and a scheduling link.