Devansh Soni

Software Engineer

Bengaluru, Karnataka, India3 yrs 10 mos experience

AI EnabledAI ML Practitioner

Key Highlights

Expert in building scalable backend systems.
Recognized with Samsung Excellence Awards.
Developed high-impact projects in Generative AI.

Stackforce AI infers this person is a Backend-focused Software Engineer with expertise in SaaS and Generative AI technologies.

Contact

sonidev0201@gmail.com LinkedIn

Skills

Core Skills

FlaskSpring BootAwsDjangoKafkaGenaiDocker

Other Skills

Agentic AILangChainMongoDBRedisgRPCKubernetesInfluxDBGoogle CloudGitHub ActionsCloud ComputingJavaMicroservicesCloud ApplicationsWeb ApplicationsDistributed Systems

About

I'm a Senior Software Engineer at Samsung R&D Institute, Bangalore, with a strong focus on building scalable and efficient backend systems. At Samsung R&D, I've played a key role in projects such as DeviceQnA for Bixby 3.0, AIOps, Generative AI platform, and several other high-impact systems, optimizing performance, scalability, and efficiency. My contributions have been recognized with Samsung Excellence Awards and the Annual MD Recognition Award. Skilled in Flask, Spring Boot, Django, Node.js, and cloud technologies like AWS, Docker, and Kubernetes, I have built robust microservices, distributed systems, and CI/CD pipelines, ensuring scalability and seamless deployments. With a strong foundation in Data Structures and Algorithms, I have solved 1,000+ problems on platforms like LeetCode, CodeChef, and GeeksforGeeks, demonstrating analytical and problem-solving skills. 📧 Email: sonidev0201@gmail.com

Experience

3 yrs 10 mos

Total Experience

2 yrs 8 mos

Average Tenure

1 yr 2 mos

Current Experience

Servicenow

Software Engineer (IC2)

Apr 2025 – Present · 1 yr 2 mos · Hyderabad, Telangana, India · Hybrid

Samsung r&d institute india

2 roles

Senior Software Engineer

Mar 2024 – Apr 2025 · 1 yr 1 mo · Bengaluru, Karnataka, India

Implemented the Backend for GenAI-based Multi-modal contextual RAG application, supporting 10+ document types, with Tabular and Image data, and using language models to generate accurate responses.
Worked on the development of Prompt studio, a platform enabling prompt development for Bixby 3, increasing efficiency by 60%, with features including prompt saving, discussion forum, and authentication.
Developed a Benchmarking tool to evaluate LLMs on inference servers e.g. TGI, Triton (with ONNX, TensorRT), etc., with/without quantization, under varying levels of concurrency, collecting metrics such as queue time, time per token, etc., to automate performance analysis.
Supported the development of UFPP (User Feedback Prediction Platform) for Bixby, achieving 95% accuracy in analyzing user sentiment and satisfaction, enhancing response quality and user experience.
Tech used: Flask, Spring Boot, GenAI, Agentic AI, LangChain, MongoDB, Redis, gRPC, AWS, Docker, Kubernetes.

FlaskSpring BootGenAIAgentic AILangChainMongoDB+5

Software Engineer

Aug 2022 – Mar 2024 · 1 yr 7 mos · Bengaluru, Karnataka, India

Enhanced scalability of Bixby Services across three locales by integrating neural-network-based models, achieving a 46% reduction in operational costs on average.
Developed a Metric Registry service to periodically retrieve data from Sumo Logic, store it in a Time-series database, and use it to predict upcoming traffic for dynamic server scaling, reducing scaling time to half.
Pioneered the creation of a Generative AI Platform, facilitating access to 20+ self-hosted Large Language Models (LLMs), empowering employees with cutting-edge AI capabilities.
Built CI/CD pipeline with Pulumi and GitHub Actions to automate AWS deployments.
Contributed to the development of a React.js front-end for monitoring predictions and identifying failures.
Tech used: Django, Spring Boot, Flask, InfluxDB, GenAI, Kafka, AWS, Google Cloud, Docker, etc.

DjangoSpring BootFlaskInfluxDBGenAIKafka+3