Ishan Nangia

Ishan Nangia

About

Detail

Building Tech for Environment@Clint 🌏
National Capital Territory of Delhi, India
Contact Ishan regarding: 
Flexible work
Starting at USD10/hour
groups
Networking

Timeline

work
Job
school
Education
folder
Project

Résumé

Jobs verified_user 0% verified
  • Clint
    Founder
    Clint
    May 2025 - Current (1 year 5 months)
    Clint Solutions is a consulting and solutions firm tackling tough challenges at the intersection of data and climate change with a heavy focus on EVs, Batteries, and Ocean-related problems.
  • iRasus Technologies
    Head of Data Science
    iRasus Technologies
    Mar 2025 - May 2025 (3 months)
    ● Designed and deployed battery and geospatial analytics algorithms (e.g., PoI detection, range estimation, SoC/SoH modelling, etc.).
    ● Created a Python CLI infrastructure that could be run as cron jobs using a SQL-based
    DB.
    ● Developed robust data cleaning and map-matching workflows for battery and GPS datasets.
    ● Managed a team of 3 interns and employees along with creation of DS strategy and roadmap
  • iRasus Technologies
    Product Manager
    iRasus Technologies
    Jul 2024 - Mar 2025 (9 months)
    ● Led end-to-end product development and roadmap execution for battery intelligence solutions, managing a 7+ cross-functional team across engineering, sales, and customer success.
    ● Oversaw a major platform migration to help resolve key engineering bottlenecks while building the go-to-market strategy.
    ● Led product pitches and technical demos for enterprise clients, managed pre-sales and post-sales efforts, and cultivated key client relationships in the mobility and battery sectors.
  • ReefBuilder
    Founder
    ReefBuilder
    May 2024 - Current (2 years 5 months)
  • Technology for Wildlife Foundation
    Computer Vision Consultant
    Technology for Wildlife Foundation
    Dec 2023 - Jun 2024 (7 months)
    ● Designed ML pipeline from collection of data to final geospatial products for restoring a degraded
    evergreen forest.
    ● Modelled and combined drone and satellite imagery to segment forest degradation indicators at scale using
    YOLO and Random Forest models.
    ● Designed and performed field work experiments to calibrate drone parameters and collect primary data for
    segmentation tasks.
    ● Used single-image photogrammetry to automate statistical measurement of objects of interest in drone
    imagery procured through a variety of DJI drones.
    ● Studied and summarised literature on ecological restoration and wildlife conservation problems.
    ● Wrote technical blogs to communicate our research and findings.
  • Coastal Impact India
    Core Team Member
    Coastal Impact India
    Oct 2023 - May 2024 (8 months)
    ● Scuba dived as a conservation diver at Grande Island to transplant corals, conduct marine surveys, and collect primary data for segmenting coral images.
    ● Setup labelling procedure for creation of datasets for coral detection
    ● Used YOLO and other foundation segmentation models to segment corals.
    ● Learnt to ID fishes, work on color correction of underwater imagery, and built 3D models of artificial
    reefs using Agisoft Metashape and Affinity.
    ● Wrote code to automate multiple tasks for Coastal Impact, organised their workspace and data documentation with Quire and Notion, and started their monthly newsletter.
  • F
    Founder
    FotoMaker
    Jan 2023 - Jun 2023 (6 months)
    FotoMaker is an online platform that allows users to create market-ready product creatives using SOTA stable diffusion-based techniques in a 3 click process.

    Working on this full-time I have at least done the following:
    ● Conducted over 80+ talks with potential users
    ● Created pitch decks, studied the market thoroughly, laid out product roadmaps, etc.
    ● Worked on frontend with ReactJS and backend with Python
    ● Set up the full machine learning architecture in a serverless manner using stable diffusion-based models
    ● Conducted extremely thorough research on stable diffusion-based models and the community around it.
    ● etc.
  • LOCUS
    Data Scientist
    LOCUS
    Dec 2021 - Nov 2022 (1 year)
    Worked on data science problems including:
    ● Sizing multiple logistics-based problems and performing time series and recommendation system-based literature surveys.
    ● Classification of Indian address using NLP-based methods
    ● Rider assignment using recommendation systems
    ● Data pipeline building
    ● Building a fully functioning rider recommendation application
  • M
    Visiting Scholar
    MPI SWS
    May 2021 - Nov 2021 (7 months)
    - Worked under Dr. Krishna Gummadi and his PHD student Till Speicher.
    - Designed an RL algorithm to make budgeted allocation of resources over time and sequential demands.
    - Studied the optimal thresholds learnt by a Dynamic Programming algorithm when considering the selection of applicants with limited slots under fairness constraints and biases in decision-making.
  • Indraprastha Institute of Information Technology Delhi
    Research Intern
    Indraprastha Institute of Information Technology Delhi
    Jan 2021 - Sep 2021 (9 months)
    - Worked under Dr. Tavpritesh Sethi to design an RL algorithm that can aid people with motor impairments.
    - Performed extensive literature survey to design the study and the data collection application.
    - Obtained IRB permissions that would allow us to collect data by leading a team of 5 researchers.
    - Wrote initial code for RL algorithm
  • Indian Institute of Technology Delhi
    Research Assistant
    Indian Institute of Technology Delhi
    Jul 2020 - Mar 2021 (9 months)
    Worked under Dr. Rijurekha Sen on the following topics:

    - Applying Differential Privacy to Gaussian Process Regression
    - Developing Sanity Checks and variants of Gaussian Process Regression models for Pollution Data collected using Mobile Sensors
    - Using Reinforcement Learning to deploy Mobile Sensors
  • Sri Venkateswara College Delhi University
    Research Assistant
    Sri Venkateswara College Delhi University
    Jan 2020 - May 2020 (5 months)
    - Worked as a RA under Dr. Veena Budhraja on DOE research - Ran python experiments to compare different types John's 3/4th fractional factorial design
  • Microsoft
    Microsoft student partner
    Microsoft
    Jan 2020 - Jan 2021 (1 year 1 month)
  • Boolean Tech Club
    President and Co-Founder
    Boolean Tech Club
    Sep 2019 - May 2020 (9 months)
    - Founded the first ever tech-society of my college - Organized cross-college/university technology focused workshops and an online tech-fest - Taught data science and python to students throughout my final semester
  • Sri Venkateswara College Delhi University
    Executive
    Sri Venkateswara College Delhi University
    Aug 2019 - Mar 2020 (8 months)
    - Worked as an executive of the Statistics Association - Responsibilities included helping organize the department fest and other events organized by the Statistics department
  • Red Dot FoundationSafecity
    Data Science Intern
    Red Dot FoundationSafecity
    Aug 2019 - Sep 2019 (2 months)
    - Built a chatbot using Dialogflow and python for Safecity's website and Facebook page - It answered common queries about the website and about Safecity - Deployed it as Flask app - Used a webhook(along with ngrock) to connect it to a database enabling it to share information on detecting the user in an emergency state
  • Genpact
    Machine Learning Intern
    Genpact
    Jun 2019 - Jul 2019 (2 months)
    - Worked on multiple NLP problems involving financial banking data - Classification of Type and Subtype of financial emails: Used multiple classification models with stratifiedkfold, randomsearchcv etc. Chose Linear_SVC as the final model and increased accuracy from 30% to 40% to 75% by experimenting with preprocessing techniques using NLTK, Gensim and Sklearn. - Classification of emails to identify which ID they were related to (with over 7000 different unique IDs) based on the payment description column: Built different similarity matrices(Jaccard, euclidean, Cosine etc.) to classify emails correctly. Designed algorithms and functions to pick up meaningful alphanumeric and numeric patterns from the data to improve accuracy of the deployed
  • Blackcoffer
    Data Scraping Intern
    Blackcoffer
    Sep 2017
    - Used BeautifulSoup to scrape and collect data from financial documents from the US SEC website.
  • I
    Founder
    Indian Beatboxers United
    Dec 2015 - Dec 2018 (3 years 1 month)
    - Founded Indian Beatboxers United to promote the Indian beatboxing scene - Organized several beatboxing competitions in India (Hard Rock Cafe Gurgaon, Delhi, Kolkata etc.) - Youtube channel was used to showcase different beatboxers from across the nation. Became the most Subscribed to Indian beatboxing channel in 2017
Education verified_user 0% verified
  • S
    Bachelors of Science : Statistics (Hons
    Sri Venkateswara College
    Jan 2016 - Dec 2020 (5 years)
    Coursework : Linear Models, Probability and Probability Distributions, Statistical Inference, Sampling Distributions, Survey Sampling Techniques, Data analysis using SPSS and R, Differential Equations, Calculus, Linear Algebra.
  • D
    High School Diploma, Commerce
    Delhi Public School Gurgaon
    Jan 2005 - Dec 2015 (11 years)
    ● I was the Music President 2014-2015
    ● Engaged in social service on several occasions
    ● Won several inter and intra school music competitions
Projects (professional or personal) verified_user 0% verified
  • I
    Internship Recommendation System with CLI
    Mar 2019 - Apr 2019 (2 months)
    I built a hybrid recommendation system which used a combination of knowledge and content based models to make recommendations and a command line interface application that can be used to scrape data from the internet and make recommendations on it.
  • R
    Recommendation system for IBM Watson Articles
    Feb 2019 - Mar 2019 (2 months)
    Analyzed the interactions that users have with articles on the IBM Watson Studio platform and made 3 recommendation models to recommend new articles to them. I used a ranking based approach, a collaborative filtering approach and a machine learning approach.
  • I
    Indian Literacy Rate Analysis
    Feb 2019 - Mar 2019 (2 months)
    I was interested in knowing more about how the literacy rate in Indian states was affected and what factors could we work on to help improve the literacy rate of different states. Using a dataset available on Kaggle and using the CRISP-DM process, I tried to answer these 3 questions in depth:

    1. Which states have the highest and the lowest literacy rates?
    2. How are the top 3 states different from the bottom 3 states and what factors can the bottom 3 states work on to increase their literacy rate?
    3. Which class is the dropout rate maximum in ?

    I have also created a medium post on this analysis that can be found here :
  • D
    Disaster Response Pipelines with Figure Eight
    Jan 2019 - Feb 2019 (2 months)
    In this project, I have applied my data engineering and NLP skills to analyze disaster data from Figure Eight to build a model for an API that classifies messages sent during a natural disaster into different categories based on what need is being highlighted in the message. I learnt how to create an ETL and a machine learning pipeline to categorize messages sent during disaster events. These maybe then directed to, according to the classified category, the appropriate disaster relief agency. The html templates were already provided by Udacity and the visualizations used have been coded by me.
  • F
    Finding Donors (Supervised)
    Dec 2018 - Jan 2019 (2 months)
    I used supervised machine learning (classification) to find people who were most likely to be donors according to several publicly available features from a dataset which originated from the UCI Machine Learning Repository. After comparing various classifiers, I performed grid search to tune the model and explored feature importance using the AdaBoostClassifier.
  • I
    Image Recognition app
    Sep 2018 - Oct 2018 (2 months)
    Built an image recognition command line app to predict the species of a flower from 102 different flowers based on the image provided by the user using pytorch in python. Got to implement transfer learning and experiment with neural networks. Also used argparse module for command line interface.
  • C
    Customer Segmentation for Mailout Campaign
    Jun 2018 - Jul 2018 (2 months)
    Worked with real-life data provided by the Bertelsmann partners AZ Direct and Arvato Finance Solution. The data here concerns a company that performs mail-order sales in Germany. Their main question of interest was to identify facets of the population that were most likely to be purchasers of their products for a mailout campaign. I used unsupervised learning techniques to organize the general population into clusters, and then used those clusters to see which of them comprised the main user base for the company. Prior to applying the machine learning methods, I also assessed and cleaned the data in order to convert the data into a usable form.