✕ CLOSE ABOUT SKILLS EXPERIENCE PROJECTS EDUCATION CONTACT

MANISH KUMAR SINGH

AI & MACHINE LEARNING ENGINEER

"Building intelligent systems that learn, adapt, and scale"

3+
END-TO-END SYSTEMS
5+
INTERNSHIPS
36K+
IMAGES TRAINED
300+
LEETCODE SOLVED
Manish Kumar Singh
01 / 07

Who I Am

// OVERVIEW
AI and Machine Learning Engineer who has built 3+ end-to-end intelligent systems, including a retinal disease detection model trained on 36,000+ images and a RAG-based LLM assistant for document QA.
// EXPERTISE
Experienced in designing scalable ML pipelines, training models on large datasets, deploying real-time inference systems, and working with LLMs, NLP, and AI-driven applications.
// LOCATION & CONTACT
📍 Bengaluru, India
📧 singhmanishofficial102@gmail.com
📞 +91 8539904102
5+
INTERNSHIPS
36K+
IMAGES TRAINED
10K+
RECORDS PROCESSED
5+
CERTIFICATIONS
02 / 07

Tech Stack

🧠
AI / ML
TensorFlowPyTorchScikit-learnNLPTransformersLLMsRAG SystemsLLM AgentsComputer Vision
💻
LANGUAGES
PythonSQLRBash
🛠️
DATA & TOOLS
LangChainFAISSPandasNumPyStreamlitFlaskOpenCVGitXGBoostTkinter
📊
DATA VISUALIZATION
MatplotlibSeabornPower BITableauEDAPredictive Analytics
🚀
DEPLOYMENT
DockerCI/CDREST APIsModel Deployment
☁️
CLOUD
AWS EC2AWS S3
PROBLEM SOLVING
LeetCode 300+Data StructuresAlgorithms
03 / 07

Work History

TECHNOHACKS SOLUTIONS PVT. LTD 2 MONTHS · REMOTE
Machine Learning Intern
REGRESSION · FLASK API · SCIKIT-LEARN
  • Built regression model to predict sales revenue, achieving strong performance (R² ≈ 0.9)
  • Developed and deployed Flask-based API for real-time inference, reducing manual workload by 40%
  • Processed and engineered features from 10,000+ records using scalable data preprocessing pipelines
  • Evaluated model performance using RMSE, error analysis, and feature importance optimization
TAKE IT SMART 4 MONTHS · REMOTE
Data Science Intern
EDA · NLP · CLASSIFICATION · REGRESSION
  • Performed exploratory data analysis and preprocessing on structured datasets to create ML-ready data
  • Built and trained machine learning models using regression and classification algorithms
  • Applied NLP techniques for text preprocessing and analysis tasks
  • Improved model performance through feature engineering and hyperparameter tuning
  • Evaluated models using accuracy, RMSE, and confusion matrix metrics
SKILLCRAFT TECHNOLOGY REMOTE
Machine Learning Intern
ML MODELS · DATA ANALYSIS · PYTHON
  • Developed and evaluated machine learning models for real-world datasets
  • Performed data preprocessing, feature selection, and model validation tasks
  • Collaborated on end-to-end ML pipeline design and documentation
HEX SOFTWARES INNOVATE REMOTE
ML / Data Science Intern
DATA SCIENCE · MODEL BUILDING · PYTHON
  • Designed and implemented data analysis workflows and predictive ML models
  • Worked on feature engineering and performance tuning for classification tasks
  • Delivered actionable insights through data visualization and reporting
CODSOFT REMOTE
AI Intern
ARTIFICIAL INTELLIGENCE · NLP · PYTHON
  • Built AI-based projects including chatbots and NLP classification systems
  • Implemented rule-based and ML-driven solutions for AI tasks
  • Gained practical experience in AI fundamentals and real-world problem framing
04 / 07

Built Systems

PROJECT · 001
Curalink — AI Medical Research Assistant
MERN + OPEN SOURCE LLM + MEDICAL DATA APIS
AI-powered system that retrieves real medical research data and generates structured insights. It combines retrieval, ranking, and reasoning to help users explore diseases, treatments, and clinical trials.
  • Fetches research papers from PubMed & OpenAlex
  • Retrieves clinical trials from ClinicalTrials.gov
  • Intelligent ranking (relevance, recency, credibility)
  • AI-generated structured summaries & insights
  • PDF report generation
  • Handles API failures with retry + fallback
Node.jsExpressReactViteHugging FacePubMed APIOpenAlex APIClinicalTrials API
VIEW FRONTEND →
VIEW BACKEND →
PROJECT · 002
LuminaPath
AI RETINAL DISEASE DETECTION
Deep learning system to detect retinal diseases from OCT images using CNN models trained on 36,000+ images. Improved accuracy from 63% to 84%.
  • Trained on Google Colab, exported weights, integrated into local backend
  • Full-stack pipeline with Streamlit frontend for real-time predictions
  • Automated PDF reports with AI-based explanations
  • Multilingual support for improved accessibility
TensorFlowPythonStreamlitFlaskCNN
VIEW ON GITHUB →
PROJECT · 003
Brainwave RAG Assistant
LLM-BASED DOCUMENT Q&A SYSTEM
Retrieval-augmented generation system for document question answering using LangChain and FAISS, generating context-aware responses for user queries.
  • Pipeline for document ingestion, chunking, embeddings, and vector search
  • Integrated LLMs for context-aware response generation
  • Improved response quality via prompt optimization and retrieval tuning
LangChainFAISSOpenAIPythonRAG
VIEW ON GITHUB →
PROJECT · 004
StockLensAI
STOCK ANALYSIS & FORECASTING
ML-based stock analysis system integrating financial datasets, FinBERT for sentiment analysis, and Prophet for time series stock trend forecasting.
  • FinBERT sentiment analysis on financial news for market signals
  • Time series forecasting using Prophet for stock trend prediction
  • Evaluated model reliability using MAPE metric
ProphetTransformersFinBERTPython
VIEW ON GITHUB →
PROJECT · 005
AI Caption & Hashtag Generator
GENERATIVE AI · NLP · CONTENT
AI-powered tool that automatically generates relevant captions and hashtags for images and content using language models and NLP techniques.
  • Leveraged LLMs for context-aware caption generation
  • Hashtag recommendation engine trained on social media patterns
  • Simple UI built with Streamlit for real-time content generation
PythonNLPLLMsStreamlit
VIEW ON GITHUB →
PROJECT · 006
Sentiment Analysis Bot
NLP · TEXT CLASSIFICATION
End-to-end sentiment analysis system that classifies text data into positive, negative, and neutral categories using NLP and transformer models.
  • Implemented text preprocessing pipeline with tokenization and cleaning
  • Fine-tuned transformer model for high-accuracy sentiment classification
  • Deployed as REST API for real-time text sentiment inference
PythonTransformersNLPFlask
VIEW ON GITHUB →
PROJECT · 007
Financial Fraud Detection
REAL-TIME FRAUD DETECTION · BANKING
High-impact ML model for detecting financial fraud in real-time banking transactions, addressing class imbalance and achieving high precision-recall performance.
  • Handled severely imbalanced datasets using SMOTE and ensemble methods
  • Achieved high AUC-ROC score for fraud vs legit transaction classification
  • Deployed as real-time inference API for banking integration
PythonScikit-learnXGBoostFlaskSMOTE
VIEW ON GITHUB →
PROJECT · 008
UPI Transaction Dashboard
DIGITAL PAYMENT ANALYTICS · INDIA
Interactive analytics dashboard analyzing UPI digital payment trends in India, providing insights into transaction volumes, patterns, and user behaviour.
  • Processed large-scale transaction datasets with Pandas and NumPy
  • Interactive visualizations using Matplotlib, Seaborn, and Tableau
  • Key business insights on payment adoption trends across demographics
PythonPandasTableauPower BISeaborn
VIEW ON GITHUB →
05 / 07

Education & Certs

EXPECTED DEC 2026
B.E. in Artificial Intelligence & Data Science
RNS Institute of Technology
BENGALURU, INDIA
Machine Learning Deep Learning Probability & Stats Algorithms
CERTIFICATIONS
☁️
Google Cloud AI Series
GOOGLE CLOUD · VERIFIED
📊
Data Cleaning & Visualization
KAGGLE · VERIFIED
🏅
Google Cloud Fundamentals
GOOGLE CLOUD · VERIFIED
🌐
Networking Fundamentals
CISCO · VERIFIED
🤖
Machine Learning A-Z
UDEMY · VERIFIED
🐍
Python for Data Science
UDEMY · VERIFIED
AI/ML Engineer Certification
AI/ML Engineer Certification
MICRO1 · APRIL 26, 2026
06 / 07

LeetCode Milestones

300+ problems solved · 50-Day & 100-Day streaks achieved

50 Day Streak 100 Day Streak
300+
PROBLEMS SOLVED
50
DAY STREAK
100
DAY STREAK
07 / 07

Let's Connect

Ready to Build Something Intelligent?

Open to internships, collaborations, and full-time ML engineering roles. Let's discuss how I can bring intelligent systems to your team.

AVAILABLE FOR OPPORTUNITIES