The catalog

370+ projects to build

Every project comes with a complete PRD, HLD & LLD. Filter by domain, difficulty or tools to find your next build.

Tip: search matches keywords anywhere — select a Domain in the filters to see exactly the projects from that field.

Request a projectCannot find what you had in mind? Tell us and we will build the spec.

6 projects

Sort

TechnologyMedium

U-Net Segmentation Dashboard for MRI Scan Review

Outlining anatomical structures and abnormalities on MRI scans is one of the most time-consuming tasks in medical imaging, often requiring slice-by-slice manual tracing by trained specialists. Research groups and imaging teams need accessible tools that generate candidate segmentations automatically so expert time is spent verifying results rather than drawing boundaries. In this project the intern builds a Streamlit web application around a U-Net model implemented with Keras and TensorFlow. Users upload a medical image, the app runs a preprocessing pipeline using OpenCV and NumPy to normalize and resize the input, and the model produces a semantic segmentation mask identifying the target regions. The intern implements clear visualization of results, showing the original scan, the predicted mask, and a blended overlay so reviewers can immediately judge segmentation quality, and uses pandas to summarize mask statistics such as segmented area for each processed image. The build gives the intern hands-on experience with the model architecture most widely used in biomedical segmentation, teaches the preprocessing discipline that medical imagery demands, and demonstrates the ability to wrap a trained network in an interface that a clinical researcher can operate without touching code.

healthcarekerasnumpyopencv+4

PRD · HLD · LLD includedView

TechnologyEasy

Optimizer Comparison Lab for SGD, Adam, and RMSProp

SGD, Adam, and RMSProp appear in every deep learning tutorial as a dropdown choice, but the differences between them are dynamic — visible only in how training actually unfolds over time. This project builds a playground for racing the optimizers against each other on equal terms. Working in Jupyter with Keras and TensorFlow, the intern trains identical copies of a simple neural network on a toy dataset, one per optimizer, and renders live loss and accuracy plots that update as training proceeds. The overlaid curves reveal each algorithm's temperament: plain SGD's steady but slow descent, RMSProp's adaptive pace, and Adam's typically rapid early progress. The intern extends the comparison by varying learning rates per optimizer, discovering that rankings between optimizers can flip when their settings change — a nuance most tutorials never mention. The project replaces cargo-cult optimizer selection with observed understanding. The intern learns what adaptive gradient methods actually adapt, how to design fair model comparisons, and how to present training dynamics visually — foundations for principled experimentation in any deep learning role.

educationresearchdata-sciencedeep-learningjupyter+3

PRD · HLD · LLD includedView

TechnologyEasy

Dropout and Weight Regularization Experiments Notebook

Overfitting is the default fate of neural networks trained on limited data, and regularization is the cure — but learners rarely get to watch the cure actually work. This project builds a notebook laboratory where dropout and weight penalties are applied, varied, and observed rather than memorized. Working in Jupyter with Keras and TensorFlow, the intern constructs a small neural network prone to overfitting by design, then runs controlled experiments: adding dropout layers at different rates, applying L1 and L2 weight regularization at different strengths, and combining the techniques. For every configuration the notebook plots training and validation loss together, so the divergence signature of overfitting — and each technique's effect in narrowing that gap — is directly visible. The experiments are organized so users can modify parameters and rerun cells to test their own hypotheses about what will happen next. The intern comes away understanding not just what regularization is but how each technique behaves and how much of it is enough, along with experience structuring reproducible notebook experiments — the same discipline that underpins credible deep learning research.

educationresearchdata-sciencedeep-learningjupyter+3

PRD · HLD · LLD includedView

TechnologyEasy

Learning Rate and Batch Size Effects Visualizer for Deep Learning

Ask a beginner why their network is not learning and the answer is usually hiding in two numbers: the learning rate and the batch size. Their effects are far easier to understand seen than described, and this project builds the tool that shows them clearly. The intern creates a visualizer, developed in Jupyter with Keras and TensorFlow, that trains a small neural network on a compact dataset repeatedly under different learning rates and batch sizes, then plots the resulting training curves together for comparison. The characteristic patterns become recognizable at a glance: divergence when the rate is too high, glacial progress when it is too low, the noisy-but-fast trajectories of small batches versus the smooth ones of large batches. Users can adjust the settings and re-run experiments quickly enough to explore the space themselves and form their own rules of thumb. The project builds diagnostic instinct — the ability to look at a loss curve and name the likely cause — which is among the most practical skills in deep learning. The intern also practices designing controlled experiments and communicating their results visually, habits that transfer directly to research and production work.

educationresearchdata-sciencedeep-learningjupyter+3

PRD · HLD · LLD includedView

TechnologyMedium

Field Crop Health Analysis via Image Segmentation

Assessing crop health across a field by eye is slow and inconsistent, and by the time stress or disease is visible during a casual walk-through, yield has often already been affected. Agronomists and agricultural researchers need a way to turn ordinary crop photographs into quantitative health measurements they can track over time. This project has the intern build a Streamlit application that segments crop imagery using two complementary approaches: classical computer vision methods, implementing Watershed and GrabCut with OpenCV, and a deep learning path using a U-Net style model built with Keras and TensorFlow. Once plants are isolated from soil and background, the app computes health indicators from the segmentation masks, such as vegetation coverage ratios and color-based stress signals derived with NumPy, and presents side-by-side visual comparisons of each technique's output so users can judge where classical methods suffice and where learned models win. By finishing the project, the intern develops a practical understanding of both traditional and neural segmentation pipelines, learns to derive domain-meaningful metrics from pixel masks rather than stopping at pretty pictures, and demonstrates the ability to package agricultural analytics into an interactive tool a non-programmer can operate.

agriculturecomputer-visionkerasnumpy+4

PRD · HLD · LLD includedView

TechnologyEasy

Train-and-Download Regression Network Builder for CSV Data

The distance between understanding neural networks and having trained one on your own data is shorter than most beginners think. This project removes the setup friction entirely: bring a CSV file, train a real network in the browser, and leave with a working model file. The intern builds a Streamlit web application where users upload a CSV dataset, pick the target column, and train a regression neural network built with Keras on TensorFlow. Training progress is visualized live through loss curves, and once training completes the app plots predictions against actual values so users can judge the fit at a glance. A download button exports the trained model, completing the cycle from raw data to portable artifact. The intern handles the practical engineering underneath: parsing uploads, basic preprocessing, choosing sensible default architectures, and keeping training responsive inside a web application. The project gives the intern end-to-end experience with the deep learning workflow — data ingestion, network construction, training visualization, and model export — while producing an approachable tool that makes a first neural network achievable for the next learner who uses it.

educationreal-estatedata-sciencedeep-learningkeras+3

PRD · HLD · LLD includedView