Kalit Inani > Work Experience

• Bedrock Data — Menlo Park, US
  Member of Technical Staff Intern
  May 2025 - Aug 2025
  

  


• Nutanix — Bengaluru, India
  Member of Technical Staff + Intern
  Jan 2022 - July 2024
  

  

  • As part of my first full-time job + internship, I contributed to the development of NDB (Nutanix Database Service), a DBaaS platform that automates and streamlines end-to-end database lifecycle management.
  • Primarly worked on the orchestration logic for provisioning, backup, and recovery services for on-premise PostgreSQL and MongoDB instances.
  • Built one-click provisioning for MongoDB Sharded Clusters, integrating Ops Manager for seamless recovery and reducing manual effort from days to minutes.
  • Enhanced security with SSL support for PostgreSQL highly available instances and MongoDB replicasets.
  • Implemented a one-click extend-storage functionality for MongoDB instances, enabling effortless scaling.
  • Domain: Software Engineering, Database Systems, Operating Systems
  • During my two-year stay in Bangalore (plus some remote work :)), I met a lot of extraordinarily knowledgable and incredibly helpful group of people!


• Central Electronics Engineering Research Institute — Pilani, India
  Research Intern
  Jan 2022 - May 2022
  Guide: Dr. Dhiraj Sangwan

  

  • As part of my research internship at CEERI, I worked on the development of a comprehensive machine learning framework for detecting and identifying drones using Radio Frequency (RF) signals from their communication with flight controllers.
  • Achieved state-of-the-art performance with 100% accuracy for drone detection and 99.73% accuracy for operational mode classification using XGBoost and Power Spectral Density (PSD) features on the DroneRF dataset.
  • Developed a signal processing pipeline implementing advanced RF feature extraction techniques including PSD, DFT, MFCC, and time-domain analysis (RMS, ZCR) for drone communication signal characterization.
  • Designed hybrid deep learning architecture combining 1D CNN feature extraction with XGBoost classification, outperforming traditional approaches and addressing class imbalance using SMOTE techniques.
  • Domain: Machine Learning, Signal Processing, Gradient Boosting Algorithms, CNNs
  • [Paper], [Code]

  

  Figure: Training the model for a 10-class classification



• JP Morgan Chase & Co. — Mumbai, India
  Quantitative Research Intern
  May 2021 - July 2021
  

  

  • As part of the Securitized Products Group (SPG), I worked on profiling and optimizing overhead function calls for cash-flow generation, reducing module execution time by 30%.
  • Conducted experiments with supervised regression analysis algorithms to develop inference models for bond pricing.
  • Domain: Algorithmic Performance, Quantitative Finance
• Indian Institute of Remote Sensing (IIRS), ISRO — Dehradun, India
  Research Intern
  May 2020 - July 2020
  Guide: Dr. P. K. Champati Ray

  

  • As part of my research internship with IIRS, I worked on land use/land cover (LULC) change detection to assess the impact of a cyclone using CNNs.
  • Developed U-Net deep learning architecture for satellite image segmentation, achieving 94% training accuracy and 78% validation accuracy on 20,000 images across 8 land cover classes using Sentinel-2 multispectral imagery.
  • Implemented comprehensive satellite data processing pipeline using Google Earth Engine, GDAL, and QGIS to extract and process multispectral bands (B2, B3, B4, B8) with automated cropping and one-hot encoding for ground truth generation from ISRO's Bhuvan platform.
  • Conducted cyclone impact assessment by analyzing land cover changes before and after Cyclone Amphan in West Bengal coastal regions, successfully identifying agricultural land flooding, river widening, and mangrove forest damage through temporal satellite image comparison.
  • Domain: Computer Vision & CNNs, Satellite Image Segmentation
  • [Report], [Code]
  • This work introduced me to practical aspects of computer science and research, showing me their potential to solve real-world problems!

  

  Figure: Segmentation of satellite imagery showing land cover changes pre/post cyclone