Sameer Kesava

Experience

• 

  Center for Cellular and Molecular Biology, Hyderabad, India

  May 2008 - Jul 2008

  Summer Research Fellow

  Modeling of chemical reaction networks in MATLAB using Flux Balance Analysis technique

• 

  University of Manitoba

  May 2009 - Jul 2009

  Summer Intern

  Research on biopolymers production from bacteria

• 

  Indian Institute of Technology, Kharagpur

  Jul 2009 - Apr 2010

  Undergraduate Student

  Production of biopolymers, polyhydroxyalkanoates (PHA), from Enterobacter cloacae IIT-BT 08 and Bacillus coagulans IIT-BT S1.

• 

  Pennsylvania State University

  Jan 2012 - Dec 2014

  Design and development of zone-annealing processing for directional crystallization of polymer/small molecule semiconductors for optimization of organic thin film solar cells and transistors.- Achieved thermal gradients greater than 70 °C/mm.- Successfully obtained highly aligned crystalline domains of C8-BTBT small molecule semiconductor for thin film transistors confirmed from grazing incidence wide-angle X-ray scattering (GIWAXS). - Design and construction of zone-annealing set-up – interfacing by LabVIEW programming and computer-aided designing in SolidWorks.- Surface and bulk characterization techniques: atomic force microscopy (AFM), near edge X-ray absorption fine structure spectroscopy (NEXAFS), X-ray diffraction (XRD) and polarized light microscopy. Examining the effects of crystallization and miscibility of polymer/fullerene semiconductor mixtures on device performance of organic photovoltaics.- Successfully deconvoluted the effect of mixed phases of polymer/fullerene mixtures on device performance using energy-filtered transmission electron microscopy (EFTEM), grazing incidence small and wide-angle X-ray scattering (GISAXS and GIWAXS), resonant soft X-ray scattering (RSOXS), and differential scanning calorimetry (DSC).- Successfully established an approach to develop structure-property relationships by comparing internal quantum efficiencies with structural characterization.- Device performance characterization from current-voltage, quantum efficiency and fluorescence quenching measurements, and diode-equation model fitting.Determination of complex refractive indices from analyses of spectroscopic ellipsometry data of thin film polymer/small molecule mixtures, block copolymers, metals and glass, and subsequent implementation in optical modeling using transfer matrix algorithm (TMA) for the determination of important optical characteristics such as number of photons absorbed in the solar cell and reflectivity of the solar cell. Show less

  • Doctoral Student

    Aug 2010 - Dec 2014

  • Teaching Assistant

    Aug 2012 - Dec 2012

  • Teaching Assistant

    Jan 2012 - May 2012
• 

  University of Oxford

  Aug 2015 - Mar 2021

  Continuation of the previous position but with additional lab managing responsibilities.Currently working on a project examining interactions between SARS-CoV-2 spike proteins and potential antibodies.Surface science expertise - thin film fabrication (vacuum deposition and solution processing), characterization, design, and process development and control for solar cell technology. Physics of light-matter interactions in such thin film systems (from monolayers to multilayers), surfaces and interfaces using polarized light in the technique called real-time or in situ spectroscopic ellipsometry complemented with X-ray scattering, electron microscopy and spectrophotometric measurement techniques.Link to publications, presentations and demos (videos): https://sites.google.com/view/svk-experience/home Show less The main research project is on examining the physics of thin film organic semiconductors (from monolayers to multilayers) in solar cells/photovoltaic devices in terms of probing the optical properties during growth under vacuum-deposition using in situ/real-time spectroscopic ellipsometry. This requires extensive data analyses using general and advanced statistical methods carried out after data acquistion by programming in Python (numpy, pandas, scipy and matplotlib libraries used for analysis and visualization). I use X-ray techniques: reflectivity, diffraction/scattering, to complement the ellipsometry analysis.Other projects have been building process control software (with GUI) for thin film deposition in PyQt, and electrical modelling and simulation (in Python) of a solar module and subsequent fabrication and testing for low-powered devices.Secondary projects have been working with collaborators for characterizing using spectroscopic ellipsometry other materials such as perovskites (for solar cells), chalcogenides, polymers, silicon, oxides (including conductive) and metals; and material-types such as monolayers, liquids and single crystals.Analytics - mathematical modelling for real-time optical monitoring (for quality control) and deciphering the optical properties of thin film systems, surfaces and interfaces: organic semiconductors (small molecules and polymers used in OLED displays and solar cells); metals, oxides (e.g. ZnO, InSnO), phase-change materials (e.g. GeSbTe: used in compact discs), perovskites and polymers. Python for scientific data analyses, visualization (including GUI programs), optical modelling and simulations (for thin film stack optimization) using libraries such as numpy, scipy, pandas, matplotlib and seaborn. Show less

  • Senior Postdoctoral Researcher

    May 2019 - Mar 2021

  • Postdoctoral Researcher

    Aug 2015 - Apr 2019
• 

  Amey

  Jul 2021 - Dec 2024

  Digital solution and full-stack software product development using data science, machine learning informed with modelling and simulation * Industries served: transport and infrastructure * Example projects: railway network modelling and simulations using physics and data science methods; LiDAR terrain model analysis; energy, carbon and cost modellingDocument analysis - information extraction and Q&A using Generative AI LLM: OpenAI GPT models with Retrieval Augmented Generation (RAG), and Optical Character RecognitionLegacy system bug fixing and upgradeStakeholder engagementDigital Twin development: transport and infrastructureTechnical and Project LeadR&DBusiness developmentBid/tender development and submissionsClient and conference presentationsFull-stack development:Backend: Python with FastAPIFrontend: React-JavaScript-TypescriptData: SQL, Azure CosmosDb, Production: Docker, Azure, Azure DevOps (for MLOps, Kanban boards, etc) Show less Digital solution and full-stack software product development using data science, machine learning informed with modelling and simulation * Industries served: transport (road and rail) and infrastructure * Example projects: data-driven traffic management tools utilising machine learning models and data science methods, economic cost modellingStakeholder engagementDigital Twin development: transport and infrastructureR&DTechnical and Project LeadClient presentationsBusiness developmentBid/tender development and submissionsFull-stack development:Backend: Python with FlaskAPIFrontend: React-JavaScriptData: SQL Production: Docker, Azure, Azure DevOps (MLOps, Kanban board, etc) Show less

  • Principal Data Scientist

    Oct 2023 - Dec 2024

  • Senior Data Scientist

    Jul 2021 - Sept 2023
• 

  Pivigo

  Aug 2021 - Sept 2021

  Data Science Fellow

  Worked as a Data Science Fellow on a 5 week intensive data science project as part of the Science to Data Science Program (https://s2ds.org/) organized and conducted by Pivigo. The clients were Electric Power Research Institute, Inc., USA (https://www.epri.com/). The project involved working in a team of 5 with big data obtained over 30 years from instrument sensors in 7 nuclear power plants (anonymised locations). The objective was to carry out a fleet-wide assessment relating sensor information with low-cycle fatigue/stress monitoring of different components with additional goals of identifying redundancies, anomalies and other general insights. The project was a business value case study with the primary aim of determining feasibility of reducing cost of plant operation, including investments for new power plants, using data science and machine learning methods. Show less