Apoorva Agrawal

Software Engineer

Bengaluru, Karnataka, India5 yrs 11 mos experience

Highly Stable

Key Highlights

Contributed to M365 Copilot at Microsoft.
Developed a graph database for Project Nalanda.
Boosted Alexa's NLU model accuracy significantly.

Stackforce AI infers this person is a Machine Learning and Software Development expert with experience in Consumer Electronics and Voice Technology.

Contact

Skills

Core Skills

Software DevelopmentMachine LearningComputer VisionData Science

Other Skills

M365 CopilotLLMsGraph databasesAzure.NETC#JavascriptDistributed systemsData validationWorkflow appsMetrics creationImage processingSemantic SegmentationApache SparkScala

About

Computer Science graduate from BITS Pilani, class of 2020. Passionate about technology for solving real-world problems. Gathering experience of building products that deliver high performance, quality, and are tuned for big data.Exploring & Learning Artificial Intelligence, Data Science, and Competitive Coding.

Experience

5 yrs 11 mos

Total Experience

5 yrs 11 mos

Average Tenure

5 yrs 11 mos

Current Experience

Microsoft

2 roles

SDE II

Sep 2022 – Present · 3 yrs 9 mos

Working on M365 Copilot, helping users onboard and efficiently use LLMs.

M365 CopilotLLMsSoftware Development

Research Fellow

Jul 2020 – Sep 2024 · 4 yrs 2 mos

Project Nalanda: Primary contributor to the project, identified and solved 2 key information discovery problems for software developers in Microsoft. Nalanda resulted in 2 internal products and a research paper.
Designed and implemented architecture of Nalanda that constituted creating and maintaining a graph database having 150M edges and 25M nodes. The pipeline to maintain this graph real-time cost less than $500 a month.
Leveraged distributed systems, Graph databases, Azure, .NET, C#, Javascript

Graph databasesAzure.NETC#JavascriptDistributed systems+1

Amazon development center

SDE Intern(Alexa Product & Tech)

Jan 2020 – Jul 2020 · 6 mos · Hyderabad Area, India

Alexa's witty skills are powered by hundreds of ML models deployed to raise Alexa’s cognitive ability and bring it closer to human. These ML models need labeled data (aka Ground Truth).
Working as part of the quality team , our goal is to build products and services which can enable voice annotation with efficiency, high quality and lowest cost
Contributed to an end to end pipeline to efficiently validate terabytes of ground truth(labelled) data which directly boosted accuracy of NLU model of Alexa.
Built workflow apps as a easy-to-use platform to be used for data labelling
Contributed in creating new metrics for accurately measuring Annotation labelling quality , implemented and successfully deployed those in the production pipeline.
Designed a novel ML model that significantly improves the quality of transcription labelling for the same business costs
Published a research paper for this work in Amazon Machine Learning Conference,2020
Learnt and demonstrated key leadership skills like Learn & Be Curious, Customer Obsession, ownership and Dive Deep!
Offered a full time role- Software Development Engineer after my internship

Machine LearningData validationWorkflow appsMetrics creation

Samsung india

Research Intern(Camera Apps-CV)

May 2019 – Jul 2019 · 2 mos · Noida Area, India

Implemented a image processing model to apply the effects of one image's segment to a target segment in any image.
Used Computer Vision techniques like Semantic Segmentation and feature extraction technique of Mange-Kantorovich theorem
Resulting model performed 150 times faster than the previous model in place and was found to be satisfactory to be rolled out to Samsung mobile devices' camera app
Offered a full time role at Samsung - Software Engineer after my internship

Computer VisionImage processingSemantic Segmentation

Couture.ai

Summer Research Intern

May 2018 – Jul 2018 · 2 mos · Bengaluru Area, India

Worked on Spark-based Implementation of User-Personalized Recommendation System using Machine Learning Algorithms.
Implemented ETL layer using Apache Spark Transformations
Reported directly to the CTO
Technologies/Tools used: Apache Spark, Scala, IntelliJ IDEA

Apache SparkScalaETLData Science