Shamik Kundu

AI Researcher

Tokyo, Japan3 yrs 6 mos experience

AI EnabledAI ML Practitioner

Key Highlights

Expert in MLOps and machine learning deployment.
Proven track record in developing fraud detection models.
Strong experience in building integrated MLOps platforms.

Stackforce AI infers this person is a Fintech-focused MLOps and Machine Learning expert.

Contact

shamikkundu13@gmail.com LinkedIn

Skills

Core Skills

MlopsData ScienceDeep LearningMachine Learning

Other Skills

KubernetesAnomaly DetectionDockerPyTorchAzureStreamlitTerraformGraph AlgorithmsTime Series AnalysisService MeshArtificial Intelligence (AI)Data AnalysisPythonRC

About

Interested in applying MLOps best practices in day-to-day ML experiments to ensure reproducibility and reduce the time taken from prototyping to production deployment.

Experience

Rakuten

Data Scientist

Jun 2022 – Present · 3 yrs 9 mos · Tokyo, Japan

1. Building an integrated MLOps platform using open source tools on self hosted kubernetes for all the data scientists in the team.
2. Integrating istio in existing applications running on private cloud.
3. Building models for fraud detection in various Rakuten financial services such as Rpay, Rakuten Card, Edy etc.

Data ScienceMachine LearningMLOpsKubernetes

Nablas

Research Engineer

Oct 2019 – May 2022 · 2 yrs 7 mos · Within 23 wards, Tokyo, Japan

1. Working on anomaly detection in image domain. Responsible for rapid prototyping and
experimenting with various state of the art supervised and unsupervised algorithms.
2. Integrated various MLOps practices in research and experiment process such as data ver-
sioning using DVC, experiment tracking using MLflow, running experiment in dockerized
environment to ensure reproducibility.
3. Co-developing a deep learning library based on PyTorch for internal company use.
Responsible for building CLI tool for the library. Making use of Docker, RabbitMQ,
Kubernetes.
4. Worked on classification of user complaints. Used various deep learning and non deep
learning algorithms. Worked on explainability of model predictions.
5. Deployed ML based webapps on Azure. Used Streamlit for webapp UI and Terraform
for creating and managing infrastructure on cloud.

Anomaly DetectionMLOpsDeep LearningDockerPyTorchAzure

Sms datatech corporation

Machine Learning Engineer

Oct 2018 – Sep 2019 · 11 mos · Within 23 wards, Tokyo, Japan

1. Worked as one of the developers in building a system for automatically colorizing Black and White Mangas. Involved in understanding and implementing various research papers to evaluate model performances and evaluating their feasibility in production system along with regular interaction with clients.
2. Worked as a sole developer in a project on finding patterns of machine failure by analyzing time series data from sensors.
3. Implemented payment system of one of the products of the organization using PAY.JP.
4. Experimented and implemented graph based algorithms on social media.

Machine LearningGraph AlgorithmsTime Series Analysis