Backend Developer + Machine Learning Engineer

Backends built
for intelligent
systems.

I’m Vaibhav. I connect the backend and the model.

I build the software that makes computer vision and machine learning usable—from image validation in a backend to web-based research tools.

View projects Download resume (PDF)
GitHub (opens in a new tab)LinkedIn (opens in a new tab)Faridabad, India
FIELD NOTES / APPLIED AI
01 / COMPUTER VISION

From image to validation.

  1. 01
    Submitted imageUser-contributed content
    INPUT
  2. 02
    Preprocess + extractPrepare image features
    DATA
  3. 03
    Detect manipulationComputer vision + machine learning
    MODEL
  4. 04
    Backend validationIntegrate the moderation result
    SYSTEM
Inside VandeMataram150

PROJECT ARCHITECTURE / SIMPLIFIED

B.Tech Computer Science · GJUST · Class of 2027Explore the work
THE THROUGHLINE

A model matters when
it becomes useful.

01 / ENGINEER

Build the backbone.

Build the backend workflows that connect inputs, processing, validation, and usable results.

02 / BUILD

Give models a purpose.

Apply computer vision, predictive modeling, or retrieval where the problem calls for it.

03 / CONNECT

Connect the system.

Connect model outputs to a backend, an analysis tool, or an interface people can use.

01 / Selected work

Built around
real problems.

From research pipelines to applied AI.
The problem, the approach, and what it became.

Further explorations

MODELS, DATA & INTERFACES
03GENAI / FINANCIAL INFORMATIONInvestaWiseMake financial information easier to navigate.LLMs · Custom retrieval

A conversational finance assistant for Indian users with different levels of financial literacy.

  1. User question
  2. Custom retrieval
  3. LLM + context
  4. Response

Problem

Investment choices and financial products can be difficult to understand without relevant context.

Approach

Connected a conversational interface to LLMs and a custom retrieval layer to make financial information easier to explore.

Outcome

A financial-guidance assistant focused on understandable responses and product discovery.

Finance assistant on GitHub (opens in a new tab)Discuss InvestaWise
04DEEP LEARNING / IMAGE CLASSIFICATIONDog Breed ClassificationAn image-classification pipeline for 120+ breeds.Python · CNN

Move from dog images to breed predictions with a complete preprocessing, training, and evaluation workflow.

  1. Image data
  2. Preprocessing
  3. CNN training
  4. Evaluation + prediction

Problem

Distinguishing dog breeds requires an image classifier that can handle visual variation across many categories.

Approach

Built a CNN-based model covering 120+ dog breeds. Optimized preprocessing and implemented Python-based performance-evaluation workflows.

Outcome

The resume reports a 25% improvement in training efficiency. Evaluation workflows were used across diverse test sets.

The 25% figure is resume-reported; its comparison baseline and measurement method are not included.

Python · CNN · Preprocessing · Model evaluationDiscuss the pipeline
05PREDICTIVE ANALYTICS / MOTORSPORTWinner Predictor — Formula 1Explore race outcomes through historical data.Azure ML · Streamlit

A predictive-analytics project using Formula 1 data from 1953–2024.

  1. Historical data
  2. Feature engineering
  3. Model training
  4. Prediction + visualization

Problem

Explore how historical race data can support predictions about Formula 1 winners.

Approach

Combined preprocessing, feature engineering, and model experimentation using Azure ML and Jupyter.

Outcome

Connected predictions and visualizations through Streamlit to explore modeled race outcomes.

Azure ML · Jupyter · StreamlitDiscuss the model
06WEB / PERSONAL SHOWCASEGet to Know MeThe interface for the work. You’re looking at its next chapter.HTML · CSS · JavaScript

Problem

Projects need clear context: what they address, how they work, and what I contributed.

Approach

A personal showcase built with semantic HTML, responsive CSS, and lightweight JavaScript. The original portfolio’s work is carried forward into this version.

Outcome

An accessible home for project stories, engineering workflows, internship experience, and my resume—with no frontend framework required.

Static site · Local fonts · Progressive enhancementA little more about me
The work continues in code.
Explore my GitHub (opens in a new tab)

02 / Experience

Learning from
the real work.

Backend integration, computer vision, and the research behind the models.

View my resume (PDF, opens in a new tab)
  1. —

    Digital India Corporation

    Noida, India

    Internee · AI image moderation

    Built the VandeMataram150 image-moderation workflow and integrated it with the production backend for real-time validation of submitted images.

    • Computer vision
    • Model evaluation
    • Backend integration
    Explore the project
  2. —

    Infosys Springboard

    Remote

    Machine Learning Intern · Virtual Internship 6.0

    Built a deep-learning image-classification pipeline spanning preprocessing, model training, evaluation, and prediction.

    • Python
    • Deep learning
    • Image classification
  3. —

    ICAR-IASRI

    New Delhi, India

    Research Intern · Agricultural computer vision

    Developed a leaf-counting pipeline with Python and OpenCV, then integrated the results into a web application for plant-phenotyping research.

    • Python
    • OpenCV
    • Plant phenotyping
    Explore the research
Vaibhav Singh standing in a white shirt with a shoulder bag
VAIBHAV SINGHBACKEND / ML

03 / Behind the work

Backend thinking.
Applied intelligence.

I’m a Backend Developer and Machine Learning Engineer, studying Computer Science at GJUST. My work connects backend software with computer vision and machine learning.

Research at ICAR-IASRI taught me to connect image analysis to a usable research tool. My time at Infosys Springboard strengthened the modeling workflow. At Digital India Corporation, I worked on bringing AI into a platform’s backend.

I care about what happens around a model: the data going in, the evaluation behind it, and the experience on the other side. I’m building deeper skills in predictive analytics, retrieval-based AI, and the computer-science fundamentals that support reliable systems.

EDUCATION / 2023–2027

B.Tech, Computer Science & Engineering

Guru Jambheshwar University of Science & Technology, Hisar

Expected graduation · June 2027

04 / Technical toolkit

The tools behind
the work.

Grouped by how I use them.
Grounded in projects and practice.

01 / LANGUAGES

Write the logic.

  • Python
  • C / C++
  • SQL
  • JavaScript
  • HTML / CSS
02 / MACHINE LEARNING

Build the model.

  • Scikit-learn
  • OpenCV
  • CNNs
  • Deep learning
  • Model evaluation
03 / DATA & ANALYTICS

Understand the inputs.

  • NumPy
  • Pandas
  • Matplotlib
  • Feature engineering
  • Jupyter
04 / APPLIED AI

Connect the intelligence.

  • Computer vision
  • LLMs
  • Custom retrieval
  • Generative AI
  • Predictive modeling
05 / BACKEND & ENGINEERING

Connect the backend.

  • Git / GitHub
  • MySQL / SQLite
  • Streamlit
  • Azure ML
  • Web interfaces
06 / FOUNDATIONS

Keep the fundamentals.

  • Data structures & algorithms
  • Object-oriented programming
  • DBMS
  • Operating systems
  • Computer networks

05 / Let’s connect

Have a useful
problem in mind?

I’m interested in early-career opportunities in backend development, machine learning, and computer vision—and conversations about practical systems.

Let’s talk
EMAIL ME DIRECTLYvaibhavsinghr01@gmail.com
LinkedIn (opens in a new tab)GitHub (opens in a new tab)

Based in Faridabad, Haryana, India.