Technical skills
- Programming languages – C, Java, Python
- Web languages – HTML, Java Script, XML, PHP, WordPress
- Databases – SQL (MySQL, PostgreSQL), NoSQL (MongoDB, Hbase, Cassandra)
- Hypervisors – Virtual Box, KVM, ESXi, Dockers
- Cloud computing tools – Cloudsim, Openstack
- Cloud services – AWS, Microsoft Azure, Google Cloud
- Big data – Hadoop (HDFS, MapReduce, YARN, Pig, Hive, Sqoop, Mahout), Spark, Kafka, Flume
- Data science – Python (Numpy, Pandas, Matplotlib, Seaborn, Plotly, Scikit-learn)
- Deep learning – Tensorflow, Keras, Pytorch
- Report writing – Latex
Hands-on experience
- Designed end-to-end digital pathology pipelines for WSI/tile classification and segmentation, including artifact filtering and tile extraction.
- Produced tumor annotations in QuPath and instituted ground-truth and QA workflows; curated a versioned, tile-level tissue dataset to enable benchmarking and reproducibility.
- Trained and deployed multimodal (image + text) pathology foundation models for automated analysis and report generation.
- Extended TensorFlow and PyTorch by overriding core modules (e.g., data loaders, schedulers, distributed hooks, inference callbacks) to customize training and inference behavior.
- Designed and deployed production-grade data analytics pipelines.
- Orchestrated GPU/HPC workloads with Slurm, optimizing scheduling, utilization, and throughput.
- Managed shared GPU servers and clusters (Sherlock, Carina) and NAS for multi-terabyte image corpora; configured GPUs for multi-user access.
- Designed and trained CNNs, RNN, encoder–decoder models, and Transformers for time-series tasks.
- Modified GANs to synthesize time-series data.
- Used SHAP to interpret predictions and quantify feature importance.
Built an anti-spoofing pipeline for facial authentication using YOLO. - Implemented reinforcement learning methods for node embedding.
- Deployed OpenStack and provisioned virtual clusters; launched Hadoop and Spark on physical and virtual environments.
- Developed distributed algorithms in Java and Python.
- Preprocessed large, unstructured datasets and built data stores using NoSQL databases.
- Implemented and benchmarked meta-heuristic optimization algorithms (ACO, ABC, GA, GE, PSO).