Saubhik Kumar

Software Engineer

Patna, Bihar, India3 yrs 7 mos experience

Most Likely To Switch

Key Highlights

Expert in Backend Development and Infrastructure Management.
Proven track record in Fintech and EdTech industries.
Strong foundation in Machine Learning and NLP.

Stackforce AI infers this person is a Backend Engineer with expertise in Fintech and EdTech sectors.

Contact

Skills

Core Skills

Backend DevelopmentInfrastructure ManagementMicroservices ArchitectureMachine Learning

Other Skills

API DevelopmentAPIsAWSAlgorithmsAmazon EC2Amazon S3CC++CI/CDCeleryClickhouseCompetitive ProgrammingComputer ScienceData StructuresDebezium

About

Hello, I am Saubhik and I am Software Engineer. Currently I am MTS 2 at Adobe. Previously I was working as a Software Engineer at Decentro, a YC20 Fintech startup. I write Backend code here, and help improve AWS infrastructure. Prior to Decentro, I worked as a Software Engineer at Unacademy, with a team dedicated to enhancing the Live Class experience for both learners and educators. I am a B.Tech. CSE graduate from IIT Patna (Batch of 2022) with CGPA of 8.56. Competencies: Python, Golang, Java, MySQL, AWS, Databases (MySQL, MongoDB, ElasticSearch, Redis, Clickhouse) My personal interests include: Tech, Sports, Film and Music. I enjoy learning new things and connecting professionally with people, so don't hesitate to reach out if you'd like to get in touch: Email: saubhik.pandey@gmail.com

Experience

Adobe

Member of Technical Staff - II

Aug 2024 – Present · 1 yr 7 mos · Noida, Uttar Pradesh, India · Hybrid

Learning how to debug a large codebase.
Learning how to fix issues with minimum context. Also learning how to explain bugs and bug fixes to customers.

Decentro

Software Engineer | Backend, Infrastructure

May 2023 – Jul 2024 · 1 yr 2 mos · Bengaluru, Karnataka, India · Hybrid

Software Engineer - wearing Backend and Infrastructure hats.
Built a realtime OLTP to OLAP replication pipeline from MySQL (AWS RDS) to Clickhouse, taking care of migration.
Developed a HULK service - for managing anything in Bulk - be it API calls or processes, reducing the API development time by 95%.
Developed APIs for Image Document Alignment and Contrast Optimization for use in OCR.
Handling multiple Infrastructure improvements projects - Kong Upgrade, Mongo Atlas to DocumentDB migration, CI/CD pipelines for all services and many more.

MySQLClickhouseAWSMongoDBCI/CDBackend Development+1

Unacademy

Software Engineer | Backend (Python, Golang)

Jun 2022 – Apr 2023 · 10 mos · Bangalore Urban, Karnataka, India · Remote

Backend Engineer at team CLX (Core Learning Experience). We build and improved features that makes Live Class experience better for Learners and Educators.
Designed and executed the backend development of a new microservice Rollout Framework, enabling seamless management of A/B rollouts for features. Integrated automatic percentage rollouts for Live classes, enhancing the feature release process.
Decoupled critical Live Class APIs from Monolith to microservice Room Service, built CRUD APIs, flow for data migration and infra creation for enhanced performance and scalability, thus reducing class load time by 80%.
Developed and integrated a doubt management feature (shortlist, un-shortlist, remove doubt) for TAs in live classes, enhancing student engagement and real-time clarification.
Built feature enabling easy recorded video upload through CSV file, implemented parsing, S3 uploading, APIs, and validation logic for Unacademy Live Class videos.
Learnt how features are built from scratch. How to create microservices. How we handle production issues. Broke production 3 times.

PythonGolangMicroservicesAPIsBackend DevelopmentMicroservices Architecture

Tata research development and design centre (trddc)

Machine Learning Intern | Python, NLP

May 2021 – Jul 2021 · 2 mos · Pune, Maharashtra, India · Remote

Proposed unsupervised method to classify Supreme Court legal case documents into 300+ subcategories, utilizing TF-IDF document similarity over legal documents and acts text to generate weak labels, resulting in efficient categorization.
Achieved average precision of 0.761 over a subset of subcategories.

PythonNLPTF-IDFMachine Learning