Profile

Computer Science undergraduate (CGPA: 9.47/10, AI & ML specialization) with strong fundamentals in Data Structures & Algorithms, Operating Systems, Computer Networks & Data Communication, and C programming. Hands-on experience building and debugging production backend systems — RESTful APIs, database-backed services, and Git-driven CI/CD deployment workflows. Microsoft Certified (AI-900, AZ-900, DP-900); hackathon finalist across Google Build With India, IEEE, and HackIndia.

Education

Chitkara University, Punjab

Bachelor of Engineering in Computer Science Engineering
2024 – 2028*

AI & ML Specialization

CGPA: 9.47/10.0

Skills
Programming Languages

Python, Java, C++, SQL, JavaScript

Machine Learning & AI

Scikit-learn, TensorFlow, Keras, binary classification, regression, feature engineering, model evaluation, experiment tracking, hyperparameter tuning

NLP

TF-IDF Vectorizer, text classification, tokenization, text preprocessing, feature engineering on unstructured text, NLTK

Data Analysis

NumPy, Pandas, Matplotlib, Seaborn, Plotly, Power BI, SciPy

Computer Vision

OpenCV, Mediapipe, AWS Rekognition, real-time object detection, gesture classification

Optimization & ML Theory

Gradient Descent(SGD, Mini-Batch), PCA, SVD, Regularization (L1/ L2), Cross-Validation, optimization algorithms

Databases

PostgreSQL, MySQL, MongoDB, SQLite

Backend & Web Development

Django, Flask, Node.js, Express.js, React (MERN Stack)

Cloud & DevOps

AWS, Microsoft Azure, Git, GitHub, Linux, Render, CI/CD Pipelines

Projects

Fake Job Posting Detector⁠

Python • Scikit-learn • TF-IDF • SMOTE • Pandas • Matplotlib • Seaborn
  • •Built a binary classification pipeline on 16,222 job posting (after deduplication) to detect fraudulent listings, addressing a 95.43% / 4.57% class imbalance using SMOTE inside a leakage-free ImbPipeline.
  • •Engineered 79 features from 18 raw columns, TF-IDF on concatenated text fields (5,000 bigram features) plus 71 structured signals including binary flags, length features, and One-Hot encoded categorical fields.
  • •Tuned LogisticRegression decision threshold via F1 sweep, improving F1 on the fraud class from 0.65(default) to 0.75 and precision from 50.8% to 76.8% while maintaining 97.8% overall accuracy.
  • •Applied 5-fold stratified cross-validation with all transformations (TF-IDF, StandardScaler, SMOTE) fit per fold, eliminating data leakage. CV F1 stable at 0.61  ± 0.03.

    • CheckInPlus⁠

    Django • PostgreSQL • AWS Rekognition • Render
  • •Deployed a cloud hosted facial recognition attendance platform handling verification for 100+ users via AWS Rekognition, reducing manual attendance logging time by ~80%.
  • •Built RESTFul APIs and role-based authentication workflows. Deployed on Render with PostgreSQL backend.
    • ISHARDEEP SINGH
    1 / 2

    FormulaFever⁠

    Python • Scikit-learn • Pandas • Matplotlib • Groq API
  • •Trained classification models on 5+ season of Formula 1 race data to predict race outcomes and driver performance rankings.
  • •Performed data preprocessing, feature engineering, and model evaluation on structured race datasets. Surfaced top 10 performance indicators via feature importance analysis.
  • •Integrated Groq API for natural language race analysis, combining ML predictions with LLM-generated insights.

    • PopShot⁠

    Python • OpenCV • MediaPipe
  • •Built a real-time gesture-controlled game achieving sub-20ms hand detection latency using MediaPipe's 21-landmark hand model and OpenCV frame processing.
  • •Applied gesture classification, object collision detection, and live score tracking at 30+ FPS on standard hardware.

    • Sportify⁠

    Flask • Django • SQLite • Bootstrap
  • •Led a 4-member team to build a full-stack e-commerce platform for sporting equipment with product catalog, cart, and other management modules.
  • •Implemented user authentication, Jinja-based frontend integration, and coordinated end-to-end feature delivery.

    • MyDiary⁠

    Django • SQLite • Bootstrap • CKEditor
  • •Developed a secure, user-authenticated digital diary application.
  • •Enforeced advanced search functionality based on date, mood, and tags.
  • •Added mood analytics and structured content management for improved user experience.
    • Professional Experience

    Web Development Intern (MERN Stack)

    Immanent Solutions
    June 2025 – July 2025 | Mohali

    • Delivered full-stack features across 3 production modules using MongoDB, Express.js, React, and Node.js.

    • Designed and shipped 10+ RESTful API endpoints with standardized response schemas, cutting frontend-backend integration time by approximately 30%.

    • Diagnosed and resolved bottleneck queries surfaced during performance debugging sessions, maintained clean version history through Git-based code review workflows.

    Leadership & Organisations

    Open Source Chandigarh

    Technical Head (ML-Ops)
    2025 – Present
  • •Leading ML-Ops initiatives for open-source projects, integrating machine learning models into CI/CD pipelines using Python and cloud platforms.
  • •Established model deployment workflows covering versioning, reproducibility, and rollback procedures for 5+ contributor teams.
  • •Mentored contributors on ML deployment best practices, experiment tracking, and collaborative development standards.

    • National Service Scheme (NSS)

    Organising Team Executive
    2025 – Present
  • •Co-organized 3 university-wide social service initiatives, coordinating logistics across 200+ volunteers.
    • Achievements
  • •Semi-Finalist, Build With India (Google) – Top 5,000 teams out of 25,000+ nationwide
  • •Semi-Finalist, Hack With Her 4.0 (IEEE) – Top 30 teams out of 700+
  • •Finalist, HackIndia 2025 – Top 50 teams out of 500
  • •Led 10+ cross-functional technical teams across academic projects and national-level hackathons.
    • Certificates
    ISHARDEEP SINGH
    2 / 2