// DATA SCIENTIST //
_
01 // ABOUT
MANJOT SINGH
DATA SCIENTIST  ·  MS APPLIED DATA SCIENCE  ·  GREEN CARD HOLDER

MS in Applied Data Science from San Jose State University. I build production ML systems — real-time pipelines, ensemble models, deep learning architectures — and I care about the whole stack: from feature engineering to Streamlit deployment. Previously at AMS Informatics where I turned data into a 40% engagement lift and a 30% efficiency gain. Actively seeking Data Scientist and Data Analyst roles.

Machine Learning Deep Learning NLP Time Series MLOps Statistical Modeling Ensemble Methods Python SQL
GITHUB LINKEDIN EMAIL RESUME
02 // SKILLS
LANGUAGES
  • Python
  • SQL
ML / DL LIBS
  • PyTorch
  • TensorFlow
  • Scikit-Learn
  • XGBoost
  • LightGBM
  • CatBoost
  • NumPy
  • Pandas
CLOUD & DATA
  • AWS S3
  • SageMaker
  • GCP
  • Snowflake
  • BigQuery
  • PostgreSQL
  • MongoDB
  • Azure
TOOLS & MLOPS
  • MLflow
  • Kafka
  • Docker
  • GitHub Actions
  • Streamlit
  • FastAPI
  • Git
CORE SKILLS
  • Feature Engineering
  • Experiment Design
  • Time Series
  • Ensemble Methods
  • SHAP / XAI
  • NLP
  • Computer Vision
VISUALIZATION
  • Matplotlib
  • Seaborn
  • Power BI
  • Gradio
03 // EXPERIENCE
AMS INFORMATICS
// JUNIOR DATA ANALYST
Sep 2022 — Oct 2023  ·  Chandigarh, India
  • Analyzed performance data across 30+ client websites to identify bounce rate and page speed trends, delivering insights that reduced average bounce rate by 22%.
  • Cross-referenced traffic, uptime, and engagement data to flag underperforming pages — directly improving average client retention by 25%.
  • Built SQL reporting dashboards tracking KPIs across 30+ sites, cutting manual reporting time by 6 hours/week and enabling faster data-driven decisions.
  • Implemented targeted content strategies using web analytics, increasing customer traffic by 40% within 6 months.
  • Consolidated data sources, reducing query times by 30% and improving system efficiency.
04 // PROJECTS
CUSTOMER CHURN PREDICTION
Ensemble ML system with 15 algorithms, 85.9% accuracy. SMOTE-Tomek rebalancing on 10K records. Live Streamlit app with real-time predictions in <200ms.
XGBoostLightGBMCatBoostStreamlitGridSearchCV
LIVE DEMO GITHUB
MARKETPULSE — STOCK PIPELINE
Real-time pipeline ingesting 10 tickers every 5min via Kafka. FinBERT sentiment + GRU/LSTM hybrid. 2.04% MAPE on AAPL. 600+ MLflow runs, weekly retraining.
KafkaFinBERTLSTMMLflowFastAPIDocker
LIVE DEMO
CONVOWEAVE-AI
Conversational AI using Mistral-7B + RoBERTa with RAG architecture. 89.3% test accuracy. 4-bit quantization cutting memory 75% with no accuracy drop.
Mistral-7BFAISSChromaDBPyTorchRAG
GITHUB
DAT.AI — MIXTURE OF EXPERTS
MoE system on GCP with specialized fine-tuned models for theory, code, and math. Improved math accuracy from 20% to 50% on the MATH benchmark.
GCPHuggingFaceFine-TuningZeroGPU
GITHUB
ANOMALY DETECTION — WSN
Anomaly detection across distributed wireless sensor node data. Statistical + ML hybrid approach for identifying network failures in real time.
PythonScikit-LearnStatistical Modeling
GITHUB
EMAIL SPAM CLASSIFIER
Multi-model spam detection comparing AdaBoost, SVM, and Stacking Classifier. NLTK preprocessing pipeline with visual performance comparison.
NLTKXGBoostScikit-LearnSeaborn
GITHUB
05 // EDUCATION
2024 — 2025
MASTER OF SCIENCE
APPLIED DATA SCIENCE
San Jose State University  ·  San Jose, CA
GPA: 3.37 / 4.0
2019 — 2023
BACHELOR OF ENGINEERING
COMPUTER SCIENCE & ENGINEERING
Chitkara University  ·  Chandigarh, India
CGPA: 9.55 / 10.0
06 // CONTACT
LET'S BUILD
SOMETHING.
Open to Data Scientist and Data Analyst roles. No sponsorship needed — Green Card holder based in Fremont, CA.
EMAIL manjot1191@gmail.com
PHONE 669-302-6269
LINKEDIN manjot-singh-6b5110228
GITHUB github.com/Manjot7
LOCATION Fremont, California