Sasya.AI & Machine Learning
Graduate student passionate about Artificial Intelligence, Machine Learning, Computer Vision, and data-driven problem solving. I build intelligent systems that create meaningful real-world impact.
A little about me
MS Software Engineering student at San José State University with a background in Electronics and Communication Engineering and experience in machine learning, computer vision, and data-driven applications. Proficient in Python, SQL, Java, MATLAB, and machine learning techniques, with hands-on experience in image processing, satellite image analysis, and predictive modeling. Seeking AI/ML internship opportunities to contribute to intelligent systems and applied research.
Academic journey
Master's, Computer Software Engineering
San José State University · San Jose, CA, USA
Bachelor of Engineering (ECE)
Stanley College of Engineering and Technology for Women · Hyderabad, India
Where I've worked
Defence Research and Development Organization (RCI)
Intern · Hyderabad, India
- Assisted in the design and simulation of Substrate Integrated Waveguide (SIW) structures using CST Studio Suite.
- Analyzed simulation results and optimized electromagnetic design parameters.
- Collaborated with engineers to validate design performance through technical analysis.
- Documented simulation findings and prepared technical reports while adhering to confidentiality requirements.
Bharat Dynamics Limited
Intern · Hyderabad, India
- Gained practical exposure to defense manufacturing processes and engineering workflows.
- Assisted engineers in technical documentation, testing, and troubleshooting activities.
- Interpreted engineering specifications and contributed to process improvement discussions.
- Strengthened analytical problem-solving and cross-functional collaboration skills.
Technical toolkit
Programming Languages
Libraries
Machine Learning
AI Domains
Tools
Featured work
RAG-Based Question Answering System
A retrieval-augmented generation service that embeds documents into a FAISS vector store and grounds LLM answers in retrieved context via a FastAPI endpoint.
- Built a retrieval augmented generation pipeline that chunks and embeds a document corpus into a FAISS vector store for semantic retrieval.
- Designed the retrieval layer with top k similarity search and prompt templates that ground answers in retrieved context, reducing hallucinated responses.
- Evaluated retrieval quality with recall@k over a held-out query set, tuning chunk size and overlap to improve answer groundedness.
- Containerized the service with Docker and served it behind a FastAPI endpoint for interactive querying.
Image Classification Pipeline with CNN Benchmarking
End-to-end image classification pipeline that benchmarks classical ML models against a CNN in TensorFlow/Keras, tracked with MLflow.
- Built an end-to-end image classification pipeline with preprocessing, normalization, and augmentation of a multi-class dataset.
- Benchmarked classical baselines (Random Forest, SVM on engineered features) against a CNN implemented in TensorFlow/Keras.
- Tracked hyperparameter sweeps in MLflow, comparing network depth, learning rate, and dropout configurations via stratified k-fold cross-validation.
- CNN outperformed the classical baselines on macro-F1; analyzed the confusion matrix to identify the most frequently misclassified classes.
Loan Default Prediction Service with Model Monitoring
Gradient-boosted and logistic regression classifiers on lending data with drift monitoring, SQL integration, and MLflow experiment tracking.
- Trained gradient boosted tree and logistic regression classifiers on a lending dataset with categorical encoding and class weighting.
- Implemented a drift monitoring routine computing population stability index and per-feature drift across time slices to flag model degradation.
- Logged runs, parameters, and metrics to MLflow for reproducible comparison across model versions.
- Improved ROC-AUC and reduced false negatives relative to the logistic regression baseline; surfaced top default drivers via feature importances.
My vision
"My long-term goal is to become an AI & Machine Learning Engineer developing intelligent, scalable, and impactful solutions — contributing to cutting-edge innovations in artificial intelligence while continuously advancing my expertise in machine learning, computer vision, and software engineering."
Proudly representing Adobe
Building community, creativity & curiosity
As an Adobe Student Ambassador, I empower fellow students to explore creative technology, design thinking, and digital storytelling. I host campus events, share Adobe tools and workflows, and build a vibrant community of aspiring creators.