Ujjwal Kumar

Lead ML Engineer

Bengaluru, Karnataka, India11 yrs 6 mos experience

Most Likely To SwitchHighly Stable

Key Highlights

Over seven years of experience in data science.
Innovative contributions to personalized offers generation.
Published research on novel conductivity sensor.

Stackforce AI infers this person is a Data Science and Machine Learning expert in the Transportation industry.

Contact

Skills

Core Skills

Machine LearningSearch And DiscoveryData Science

Other Skills

AWSApache SparkCC++Data Structure and AlgorithmsForecastingGitIndian Classical MusicJavaKerasMatlabPythonRReinforcement LearningScala

About

With over seven years of experience in data science and machine learning, I am a passionate and innovative Sr. Software Engineer - Machine Learning at LinkedIn, where I work on search and discovery for LinkedIn Learning. I have a strong background in applying machine learning techniques to various domains, such as customer experience, forecasting, attribution, and personalized offers. I have also obtained multiple certifications from Coursera, such as Machine Learning, R Programming, and Exploratory Data Analysis, to enhance my skills and knowledge. I am proficient in Python, Scala, and other programming languages, as well as tools like Apache Spark, Hive, Keras, and AWS. I have also contributed to the research and development of systems and methods for scalable, adaptive, real-time personalized offers generation, which resulted in a patent. Additionally, I have published a paper on a novel conductivity sensor for measuring solution conductivity in a symposium. My core competencies include data analysis, machine learning, programming, problem-solving, and innovation. My mission is to leverage my expertise in search and discovery to create impactful and engaging learning experiences for LinkedIn users. I value collaboration, learning, and excellence, and I am eager to bring diverse perspectives and experiences to the team.

Experience

2 roles

Sr. Applied Research Engineer- Machine Learning

Promoted

Oct 2021 – Present · 4 yrs 5 mos

Reinforcement Learning Platform - Talent Marketplace Data Foundations
Member and Jobs embeddings platform, LinkedIn Talent Solutions
Search and discovery, LinkedIn Learning

Reinforcement LearningSearch and DiscoveryMachine Learning

Applied Research Engineer - Machine Learning

Oct 2019 – Sep 2021 · 1 yr 11 mos

- Search and discovery, LinkedIn Learning.

Search and DiscoveryMachine Learning

Ola (ani technologies pvt. ltd)

2 roles

Data Scientist II

Promoted

Apr 2017 – Oct 2019 · 2 yrs 6 mos

Productionised improvements to existing ML system to suggest pickup-points for 50M+ Ola users worldwide; implemented using Apache Spark in Scala & Python.
Leading projects in Customer experience and CLM, working both as a mentor and individual contributor.
Long term research project, explored mix of classical time-series and LSTM
models, usage of additional signals and data preparation steps; improved
forecasting error across models for tactical and strategic use-cases.
Implemented using Hive for data querying; Apache Spark, Pandas and Numpy
to pre-process data and Keras to train LSTM models on AWS GPU instances.
Developed attribution models to calculate experiment impacts on multiple key metrics across cities.

Apache SparkScalaPythonData ScienceForecastingKeras+2

Data Scientist - I

Oct 2016 – Mar 2017 · 5 mos

Developed demand forecasting models in R, with real-time correction to address unusual days and achieved upto 30% improvement in forecasting error.
Simulated Ola peak-pricing system, to identify unusual pricing cases and quick fixes to minimise revenue loss.

RForecastingData ScienceMachine Learning

Fico®

2 roles

Analytic Science - Scientist II

Promoted

Dec 2015 – Oct 2016 · 10 mos

Developed an ALS based recommendation model using MLLib/Apache Spark in Java.
Implemented ALS from scratch in R, with a custom cost function and shell scripts to parallelise model training.
Integrated model with FICO’s Streaming Analytics solution using its Java SDK.

Apache SparkJavaMachine LearningData Science

Analytic Science - Scientist I

Jun 2014 – Nov 2015 · 1 yr 5 mos

Implemented TBM(Temporal Behavioural Maps, a fast data processing and feature generation technique) from scratch in Java, ensuring extensibility and recursive application of feature-generation templates.
Designed and implemented a relatively lite-weight clustering based matched-sampling approach in R for causal inferencing.
Re-trained Time-to-event models, proactively identified, discussed and implemented improvements in the existing platform.

JavaRData ScienceMachine Learning

Shree cement ltd

Summer Intern

May 2013 – Jul 2013 · 2 mos · Beawar, Rajasthan

1. Designed automation solutions for manual operations and modifications in automated sections under cost-reliability tradeoffs.
2. Implemented a prioritized process parameters’ list out of over 10,000 plant-parameters; with process, operation & instrumentation perspective.
3. Suggested better measuring instrument options in terms of accuracy and compatibility for critical process parameters.