Anju Vilashni Nandhakumar

AI Engineer · Computer Vision & Reinforcement Learning

Designing interpretable and human-centered intelligent systems.

📍 Boston, MA · Open to full-time AI roles

About Me

I’m an AI Engineer with a Master’s in Artificial Intelligence from Northeastern University, passionate about building intelligent systems that bridge human intuition and machine reasoning.

My work explores the intersection of computer vision, reinforcement learning, and explainable AI — designing models that not only perform well but can be trusted and understood. I believe AI should feel less like a black box and more like a dialogue — between logic and empathy, data and design.

🧠
AI Research

Exploring interpretable systems in vision and healthcare.

🎓
Education

M.S. Artificial Intelligence – Northeastern University

💡
Innovation

Integrating NLP and CV for accessible, explainable tools.

🤝
Mentorship

Guiding learners in ML and AI-driven projects.

Experience

3+ years building production ML systems across healthcare, recruiting, and nonprofit sectors

Machine Learning Engineer (Volunteer)

Community Dreams Foundation

Sep 2025 – Present Boston, MA
  • Engineered end-to-end ML pipelines using Python, TensorFlow, and PyTorch, achieving 87% average accuracy for donor prediction while reducing inference latency by 20%
  • Collaborated with cross-functional leadership to integrate ML solutions into 3 operational programs, improving decision-making efficiency by 30% through predictive analytics
Python TensorFlow PyTorch Predictive Analytics MLOps

AI/ML Technical Lead

Design and Innovation Club

May 2022 – May 2023 Chennai, India
  • Led development of production ML systems including NLP-powered job search platform and CNN-based medical imaging tool for pneumonia detection
  • Mentored 50+ students in applied ML using Python, TensorFlow, and OpenCV, conducting hands-on workshops on model deployment
  • Built and deployed ML models using Docker containers and Flask APIs, implementing CI/CD practices for continuous model improvement
Python TensorFlow Docker Flask Mentorship

Machine Learning Intern

Jobdae

Oct 2020 – Feb 2021 Bangalore, India
  • Improved candidate-job match accuracy by 40% through Deep Learning ranking models and Predictive Analytics, implementing A/B Testing and model optimization techniques
  • Built sentiment-driven recommendation engine using Natural Language Processing, increasing placement conversions by 25% and user engagement by 40%
Deep Learning NLP Recommendation Systems A/B Testing

Projects

Computer Vision

Tumor Classification

Explainable AI for Tumor Classification

Multi-class brain tumor classifier with Grad-CAM explanations for transparent medical decision-making.

Computer VisionGrad-CAMHealthcare AI
Deepfake Detection

VisAIble – Deepfake Detection

Human-AI collaborative deepfake detection using CNN classifiers and visual interpretability tools.

Deep LearningExplainabilityResNet
Indian Sign Language Recognition

Indian Sign Language Recognition

Real-time gesture recognition system translating ISL signs using CNN + live video inference.

Real-time CVAccessibilityDetection
Music Recommendation

Emotion-Based Music Recommendation

Facial-emotion-driven music recommender using CNN emotion predictions + mood mapping logic.

Emotion AIRecommendationCV
Chest X-Ray Classification

Explainable Chest X-Ray Classification

Clinical-grade pneumonia detection with Grad-CAM heatmaps, automated PDF reporting, and patient archiving.

Medical ImagingExplainabilityComputer Vision

Natural Language Processing

Logical Fallacy Detection

Logical Fallacy Detection

Transformer-based classifier using case-based reasoning to detect logical fallacies in arguments.

NLPTransformersExplainability
Lyrics Classifier

150K Lyrics Genre Classifier

Large-scale lyrics genre prediction using transformer text embeddings on 150K song dataset.

Text ClassificationNLPTransformers

Reinforcement Learning

Parking Management RL

Multi-Agent RL for Parking Optimization

Multi-agent system that learns optimal parking allocation and traffic flow strategies.

Reinforcement LearningMulti-AgentOptimization
Highway RL

Highway Reinforcement Learning

Agent trained in highway-env for safe lane changes and strategic driving using advanced RL algorithms.

AutonomySafetyRL

Technical Expertise

3+ years deploying production AI systems with 95%+ accuracy and 40% performance improvements

🔬

Computer Vision & Deep Learning

CNNs, ResNet, EfficientNet for medical imaging, deepfake detection, and real-time classification. Achieved 95%+ accuracy with Grad-CAM explainability.

TensorFlow PyTorch Keras OpenCV
💬

Natural Language Processing

Sentiment analysis, text classification on 150K+ datasets, transformer architectures. Built recommendation engines with 40% engagement boost.

Hugging Face Transformers NLTK spaCy
🎯

Reinforcement Learning

Rainbow DQN, A3C, Decision Transformers for autonomous navigation. Multi-agent systems with 48.2 max reward in highway environments.

Gym Highway-env A3C Rainbow DQN
⚙️

MLOps & Production Deployment

End-to-end ML pipelines with Docker, AWS (EC2, S3, SageMaker), CI/CD workflows. Reduced inference latency by 20% through GPU acceleration.

Docker Kubernetes AWS Streamlit
🔍

Explainable AI & Interpretability

Grad-CAM, SHAP, LIME for transparent AI decision-making. Specialized in healthcare and cybersecurity applications.

Grad-CAM SHAP LIME
Full Tech Stack

ML/DL: TensorFlow · PyTorch · Keras · Scikit-learn · OpenCV · Pandas · NumPy

NLP: Hugging Face · NLTK · spaCy · Transformers

RL: Gym · Stable Baselines · Highway-env · A3C · Rainbow DQN · Decision Transformers

XAI: Grad-CAM · SHAP · LIME

MLOps: Docker · Kubernetes · AWS · Git · Streamlit · Flask · REST APIs · CI/CD

Publications

📄

A Deep Learning Approach to Music Recommendation based on Facial Emotion Recognition

International Journal of Research and Analytical Reviews (IJRAR)

December 10, 2022

Developed a novel approach combining computer vision and recommendation systems to predict music preferences based on real-time facial emotion analysis.

View Publication

Let's Connect

Currently seeking full-time ML Engineer roles in Boston or remote. Open to Computer Vision, NLP, Healthcare AI, and Applied ML opportunities.