Parijat Mazumdar

Director of Engineering

San Francisco, California, United States10 yrs 5 mos experience

Most Likely To SwitchHighly Stable

Key Highlights

Stackforce AI infers this person is a SaaS-focused engineering leader with expertise in machine learning and distributed systems.

Machine LearningDistributed Systems

PythonGoC++JavaCassandraRedisTornadoReact.jsElasticSearchApache FlinkData Science

10 yrs 5 mos

Total Experience

3 yrs 5 mos

Average Tenure

9 yrs 11 mos

Current Experience

Jun 2016 – Present · 9 yrs 11 mos · San Francisco Bay Area · Hybrid

Engineering Lead for Maps Optimization. We drive business growth by ensuring users' geospatial interactions are seamless, intelligent, and optimized for every ride and delivery. My organization comprises of 5 teams across 4 core areas:
1. Maps Search: Provide seamless, insightful location search and recommendation experience to riders, eaters & earners.
2. Routing: Find optimal path between locations considering travel time, distance, safety and earner/rider preferences.
3. ETAs: Provide accurate travel time predictions between locations.
4. Waypoint Selection: Recommend optimal access points (lat/lng) for rides and delivery pickup & dropoff.
Previously:
1. Tech Lead Manager, Matching: Built and scaled out the foundational matching tech supporting Uber's schedule ahead products (reserved rides, scheduled delivery orders etc). By the end of my tenure, I was the technical lead for a group of ~20 engineers and directly managed a team of 6 engineers owning the recommendation system powering earner-trip matching.
2. Software Engineer, Driver Positioning & Incentives: Built out several products providing various forms of positioning recommendations to drivers in between trips. Back in 2017-2018, I built a new high scale service from scratch which still continues to be the core service serving positioning incentives to Uber drivers in realtime.

Machine LearningPythonGoC++JavaDistributed Systems+6

May 2015 – Jul 2015 · 2 mos · San Francisco Bay Area

Data ScienceMachine Learning

May 2014 – Aug 2014 · 3 mos

Shogun Machine Learning Toolbox (http://www.shogun-toolbox.org/) is an open-source machine learning library implemented in C++ with bindings for all major languages. As part of Google Summer of Code program, I added a variety of machine learning algorithms to the library:
KMeans++
Decision tree algorithms (CART, C4.5, CHAID)
Random Forest
Stochastic gradient boosting
fast NN search using KD-trees and Ball trees
Kernel density estimation
Expectation Maximization and mixture models.
I also created some ipython notebooks as tutorials for new users:
Kernel Density Estimation: http://www.shogun-toolbox.org/notebook/latest/KernelDensity.html
Decision Trees: http://www.shogun-toolbox.org/notebook/latest/DecisionTrees.html
Random Forest: http://www.shogun-toolbox.org/notebook/latest/TreeEnsemble.html
KMeans: http://www.shogun-toolbox.org/notebook/latest/KMeans.html