Kumuda Priya Sanka

Software Engineer

Bengaluru, Karnataka, India7 yrs 9 mos experience

Highly Stable

Key Highlights

6 years of experience in Computer Vision and Machine Learning.
Mentored interns into full-time roles successfully.
Hosted award-winning Snowflake integration project.

Stackforce AI infers this person is a SaaS-focused Software Engineer with expertise in Machine Learning and Computer Vision.

Contact

kumudapriya.sanka@gmail.com LinkedIn

Skills

Core Skills

Computer VisionMachine LearningBig DataDeep Learning

Other Skills

Problem SolvingNatural Language Processing (NLP)PythonData PrivacyCloud ComputingREST APIsJavaDesign PatternsSOLID Design PrinciplesC++AlgorithmsData StructuresSoftware Development

About

6 years of experience in Computer Vision, Neural Networks, LLMs, Big Data Pipelines, and Data Privacy. I’m passionate about solving complex challenges and delivering scalable, high-performance solutions. I also enjoy mentoring and have successfully guided interns toward full-time roles, including hosting award-winning Snowflake integration project that was demonstrated at Cloud Next 2023.

Experience

7 yrs 9 mos

Total Experience

4 yrs 6 mos

Average Tenure

3 yrs 3 mos

Current Experience

Google

Software Engineer

Jan 2023 – Present · 3 yrs 3 mos · Pune, Maharashtra, India

Finetuned and Distilled a Bert-Like LLM to detect Date of Birth in text, currently handling an average of 350M requests per day. You can explore the DOB detection model here: Try it out https://cloud.google.com/dlp/demo/#!/
Developed large-scale parallel data processing pipelines using Flume(google internal version of Apache Beam) for evaluating LLMs.
Contributed to a backend service that provides a unified API for accessing multi-cloud, structured, and unstructured data. Designed the API to fetch structured data, specifically focusing on AWS Redshift.
Mentored interns, helping them successfully convert to full-time roles. Hosted an intern whose Snowflake integration project with Cloud DLP won Best Intern Project in Pune and was showcased at Cloud Next 2023. Watch here https://youtu.be/f0RLO1QSp8c?t=1237

Problem SolvingComputer VisionMachine Learning

Samsung electronics

4 roles

Associate Staff engineer, ML Solution

Promoted

Mar 2021 – Jan 2023 · 1 yr 10 mos

Optimizing Compiler Team for Deep Learning. The team is responsible for performance of deep-learning neural networks on NPU (built into mobile phones)
Responsibilities:
Design Optimizations related to runtime latency of the Neural Networks
Design Operator-level optimizations
Achievements: Awarded Best Employee Award multiple times, appreciating my contributions
Important Contributions: Optimized the memory transfer scheduling algorithm. This has shown an improved performance of 4 % on average for all the use-cases.

Problem SolvingDeep Learning

Software Engineer, ML Solution

Apr 2020 – Feb 2021 · 10 mos

Problem Solving

Software Engineer, ADAS

Jun 2018 – Apr 2020 · 1 yr 10 mos

Worked on SIMD Computer Vision Algorithms for DSP in Samsung Automotive SoC
Implemented Fixed Point implementation for RGB to LAB Color Conversion using 256-sized 16-bit LUT, Image Filtering, Gaussian Blur, SoftMax, Sparse and Dense Optical Flow

Problem Solving

Student Trainee

May 2017 – Jun 2017 · 1 mo · Bengaluru Area, India

Developed a real time engine POC for categorising and analysing bugs in the mobile broadband defect logs. (Team of 2)
Received the Best Intern Software Project for the summer of 2017.

Problem Solving