RESEARCH • ENGINEERING • INNOVATION

Engineering Intelligence

AI-ML & Computer Vision Mobile AI Applications Remote Sensing & Scientific Computing

Hi there!, I'm Uvesh Mansuri. AI-ML Engineer and SAC ISRO Apprentice working on AI&ML from satellite data pipelines to production-level smart AI applications to computer vision applications. Previously worked at Nova Softwares.

Uvesh Mansuri

About Me

AI Engineer bridging scientific research and intelligent applications.

The Vision

At the moment, I work as an Apprentice at Space Applications Centre (ISRO), Ahmedabad, India where I work on satellite rainfall validation using INSAT-3DS and NASA IMERG data through scientific Python pipelines.


Previously, I have worked on developing production AI applications as a software developer at Nova Softwares, where I developed face recognition and anti-spoofing in real-time for mobile application, computer vision pipelines and Gen AI modules, and enterprise Mobile and AI application modules. In addition, I am also enrolled in the MCA Manipal University Jaipur (Online) where I have scored 8.70 SGPA in the first semester.

My interest lies in the field of AI engineering, scientific computing, and mobile application development that involve intelligent systems solving practical problems through research and industry collaboration.

Uvesh Mansuri Professional
0
Satellite Samples Processed
0
Production Applications

My Mission

Building intelligent systems that bridge AI research and real-world applications from validating satellite rainfall data at ISRO to deploying offline face recognition on mobile devices. I aim to push the boundaries of what's possible with scientific computing, computer vision, and smart mobile AI Applications.

0
Face Recognition Accuracy
0
ML Pipeline Optimization
0
Computer Vision Systems
0
Years Professional Experience

Professional Experience

From scientific research at ISRO to production AI systems.

ISRO Logo

Space Applications Centre (ISRO)

Apprentice Scientific Computing & Remote Sensing
Apr 2026 — Present
  • Building machine learning solutions for satellite-based rainfall estimation using INSAT-3DS and NASA IMERG datasets.
  • Designing Python-based automatic pipelines for working with NetCDF and HDF5-based geospatial big data.
  • Validation of satellite rainfall predictions through statistical methods and geospatial comparison approaches.
  • Generating data visualizations and geospatial maps for scientific research purposes.
  • Utilization of remote sensing and data science in solving real-world problems.
INSAT-3DS NASA IMERG NetCDF HDF5 Python Automation Spatial Regridding Statistical Analysis Scientific Visualization Random Forest Xarray
Nova Softwares Logo

Nova Softwares

AI Application Developer
Aug 2025 — Apr 2026
  • Built production AI applications using Flutter & Python.
  • Developed offline Face Recognition & Liveness Detection with Geo-Fencing for Android & iOS.
  • Optimized TensorFlow Lite inference for low-end devices.
  • Created NLP-powered ERP Assistant for enterprise search.
  • Developed CCTV analytics using YOLO and OpenCV.
  • Implemented REST APIs and other backend integrations.
  • Delivered scalable AI solutions with cross-functional teams.
Face Recognition TensorFlow Lite Flutter YOLO ERP Assistant REST APIs SQL Chatbot Gen AI

Career Journey

From computer science student to AI engineer at ISRO.

2022

Started BCA

Began Bachelor of Computer Applications at Narmada College of Science & Commerce (NCSC), Bharuch.

2024

Application Development & Projects

Started learning app development. Created Coding Hub mobile app. Began Flutter development and worked on NCSC College App.

2025

BCA Completed • Nova Softwares

Completed BCA with 8.67 CGPA. Started Pursuing MCA at Manipal University Jaipur (Online). Joined Nova Softwares as AI Application Developer built face recognition, YOLO integration, ERP systems.

2026

Space Applications Centre – ISRO

Apprentice at SAC (ISRO), Ahmedabad satellite rainfall validation, scientific Python, remote sensing. Scored 8.70 SGPA in MCA 1st Semester.

Future

Mobile & AI-ML Engineer

Continuing to build intelligent systems at the intersection of AI, scientific research, and real-world applications.

Expertise Matrix

Technologies and tools powering research and production systems.

Programming Languages

PythonPython JavaJava C++C++ KotlinKotlin DartDart JSJavaScript SQL

AI & Machine Learning

OpenCVOpenCV YOLO Computer Vision TFLiteTensorFlow Lite Generative AI NLP LLMs scikit-learn

Scientific Computing

NumPyNumPy PandasPandas Xarray MatplotlibMatplotlib Cartopy NetCDF HDF5 Statistical Analysis Remote Sensing

Backend & APIs

FirebaseFirebase FastAPIFastAPI FlaskFlask REST APIs WebSockets

Mobile Development

FlutterFlutter AndroidAndroid FirebaseFirebase SQLiteSQLite

Databases

Firebase Realtime DB Firestore MSSQLMSSQL SQLiteSQLite

Tools & Platforms

Android StudioAndroid Studio GitGit GitHubGitHub PostmanPostman VS CodeVS Code PyCharmPyCharm

Technology Arsenal

The tools powering intelligence, research, and production systems.

PythonPython FlutterFlutter TensorFlowTensorFlow OpenCVOpenCV FastAPIFastAPI FirebaseFirebase GitGit
NumPyNumPy PandasPandas YOLOYOLO AndroidAndroid DartDart MatplotlibMatplotlib FlaskFlask

Research Work

On-going Scientific investigations at ISRO Space Applications Centre.

INSAT-3DS Monsoon Rainfall Performance Evaluation

Evaluated the performance of INSAT-3DS daily precipitation products (L3B_HEM, L3B_IMC, L3G_IMR, and L3G_GPI) during the 2025 Indian Summer Monsoon (June–September) using GPM IMERG Final Run as the benchmark dataset. Developed automated Python workflows for pixel-level comparison, spatial regridding, and temporal aggregation over the Indian domain (0°–40°N, 60°–100°E).

INSAT-3DS NASA IMERG GPM Final Run Monsoon Rainfall Evaluation Pixel-Level Validation

Processing Multi-Sources Satellite Rainfall Datasets

Performed spatial interpolation/regridding of processed multi-source satellite precipitation datasets on heterogeneous grids, daily to seasonal rainfall aggregation, and monsoon accumulation study. Experienced with NetCDF/HDF5 data types and Xarray for multidimensional geospatial analysis.

Spatial Regridding Temporal Aggregation NetCDF/HDF5/Xarray

Evaluation Based on Statistical and Intensity Metrics

Evaluated the performance of rainfall retrievals with continuous metrics like Bias, RMSE, MAE, and Correlation Coefficient. Evaluated rainfall retrievals through intensity-based metrics by classifying them based on rainfall intensity (light, moderate, heavy, and extreme) and calculating metrics like Probability of Detection (POD), False Alarm Ratio (FAR), and Critical Success Index (CSI).

RMSE • MAE • Correlation POD • FAR • CSI Rainfall Intensity Assessment

Development of Automated Scientific Python Pipeline

Developed end-to-end automated scientific pipeline using Python, NumPy, Pandas, Xarray, Matplotlib, Cartopy, and scikit-learn to process satellite data efficiently. Produced publication quality figures viz. seasonal rainfall distributions, rainfall time series, accumulation maps, and statistical comparison figures.

Python Automation Scientific Computing

Selected Works

A showcase of AI, mobile, and computer vision innovation.

NCSC College App — Smart Campus ERP Platform
Flutter Firebase Python – OpenCV Gemini API Cross-Platform

NCSC College App

A smart campus platform with role-based access for admins, faculty, staff and students. It uses real-time face recognition for attendance, AI-powered timetable scheduling, and automatic test generation using Gemini API. Interactive dashboards track marks, assignments, attendance, and announcements, improving efficiency, decision-making, communication, and overall academic management for modern institutions daily.

Multi-Role ERP Face Attendance AI Test Generation
Coding Hub — Learning Platform
Android Firebase Auth Firebase Realtime Database FCM Rest API

Coding Hub Learning

A modern learning platform featuring AI-powered quiz generation with Gemini API, interactive peer discussion forums inspired by Stack Overflow, and daily tech news updates. It helps users test knowledge, solve doubts collaboratively, stay updated with industry trends, and build skills through engaging, smart, and community-driven learning experiences every day.

Discussion Forums Tech News Community Learning
AI Liveness Detection — YOLO Anti-Spoofing Model
Python YOLO Deep Learning Computer Vision

AI Liveness Detection Model

A standalone YOLO-based real-time anti-spoofing model trained on 4750+ images to distinguish between live faces and spoofed attempts (photos, screens, masks). Built as an independent Python deep learning project for face liveness verification research.

4750+ Training Images Real-Time Detection Anti-Spoofing
Notify — AI-Powered Smart Notes
Flutter Gemini API Vector DB AI

Notify: AI Smart Notes

An intelligent notes application leveraging vector embeddings and Google Gemini AI for semantic search and contextual retrieval. Notes are stored with vector representations enabling smart similarity-based discovery across your knowledge base.

Vector Embeddings Semantic Search Gemini AI

Key Highlights

Milestones from research, engineering, and production systems.

ISRO Apprentice

Working at India's premier space research organization on satellite data science.

Scientific Data Processing

Processing 4000+ satellite samples with automated Python pipelines.

Production AI Applications

Shipped 5+ smart applications with AI-powered features to production.

Offline Face Recognition

Built offline face recognition with anti-spoofing running at 15–30 FPS.

Computer Vision Systems

Real-time YOLO detection and TensorFlow Lite inference on mobile devices.

Enterprise AI Solutions

SQL chatbot, ERP assistant, and intelligent automation for businesses.

Get in Touch

Available for AI Engineering, Computer Vision, Remote Sensing, Research & Scientific Computing roles.

Bot Avatar
Uvesh's AI Agent
Hi! I'm Uvesh's AI agent. Ask me about his projects or tech stack!