Prahalad Talur

prahalad.dev

Engineering reliable systems from sensor data.

Rising senior at Eastlake High School focused on mechanical engineering, embedded systems, aerospace, and applied machine learning.

View research repo LinkedIn

I build training pipelines,
model comparisons,
reproducible benchmarks,
and clear evidence for engineering decisions.

Featured project

N-CMAPSS Engine RUL Prediction

A research-grade deep learning pipeline for predicting remaining useful life in NASA turbofan engine data. The repo is organized around one data loader, one training CLI, a model registry, RUL-specific metrics, and Weights & Biases logging for serious benchmark runs.

Python TensorFlow Time series Predictive maintenance Aerospace
Pipeline overview for the N-CMAPSS RUL prediction project
Dataset loading, sequence preparation, training, and benchmark reporting pipeline.

Benchmark evidence

Controlled comparisons, not one-off claims.

6.495 RMSE on FD2 with MSTCN in the controlled apples-to-apples benchmark.
99.01% Accuracy@20 on FD2 for MSTCN under the same benchmark protocol.
22 Registered neural architectures across recurrent, convolutional, hybrid, and transformer families.
4.8% Gap from a verified published N-CMAPSS DS02 RMSE reference in the repo's paper-safe summary.
FD1 architecture leaderboard chart
FD1 architecture leaderboard from the controlled benchmark set.
MSTCN model architecture diagram
MSTCN architecture: multi-scale temporal convolution with fusion attention.
Comparison with external N-CMAPSS references
External N-CMAPSS comparison used to frame paper-safe result claims.

Activity

STEM, finance, and team competition.

National Personal Finance Challenge

Eastlake High School team advanced to the Round of 16 at the 2026 National Personal Finance Challenge, finishing in the top 10 out of 7,000+ teams and 20,000 students nationwide.

Research and engineering focus

Working at the intersection of mechanical engineering, aerospace, embedded systems, and machine learning, with current depth in engine health modeling and reproducible ML evaluation.

Computer Science Honor Society

Connected with Eastlake's computer science community while building software projects that connect classroom work to engineering problems.

Contact

Building toward aerospace and engineering research.

The best starting points are the research repository and LinkedIn profile.

GitHub LinkedIn