Professional Context
Balancing the need for meticulous evidence analysis with the pressure to meet tight deadlines, Forensic Science Technicians must navigate a delicate dance between quality assurance and time-to-completion, all while minimizing error rates that can have severe consequences in the pursuit of justice. This tension is ever-present, as technicians strive to deliver accurate and reliable results within the constraints of their workload and available resources.
💡 Expert Advice & Considerations
Don't rely on Gemini to replace your own expertise; use it to augment your data interpretation skills, especially when dealing with complex Google ecosystem workflows.
Advanced Prompt Library
4 Expert PromptsEvidential Hair Sample Comparison
Given a dataset of 100 hair samples from a crime scene, with each sample described by 20 morphological features, and a reference dataset of 500 known hair samples from various individuals, use Google Cloud's AutoML to develop a classification model that can distinguish between human and animal hair, and then apply this model to the crime scene samples to identify potential human hair evidence, providing a ranked list of the top 10 most likely human hair samples along with their corresponding probability scores and feature importance values.
DNA Profile Reconstruction from Fragmented Data
Using the Google Genomics API, reconstruct a DNA profile from a set of fragmented DNA sequences, each 100-200 base pairs in length, by first aligning the sequences to a reference genome using the Burrows-Wheeler Aligner, then applying a Bayesian inference algorithm to infer the most likely complete DNA profile, considering a set of 10 known alleles for each of the 20 CODIS loci, and finally evaluating the reconstructed profile against a database of 10,000 known DNA profiles to identify potential matches, reporting the top 5 matches along with their corresponding match probabilities.
Gunshot Residue Particle Analysis
Analyze a set of scanning electron microscopy (SEM) images of gunshot residue (GSR) particles, each image containing approximately 100 particles, to classify the particles into one of three categories (lead, barium, or antimony) based on their elemental composition, using a Google Cloud Vision API-based approach that applies a pre-trained convolutional neural network (CNN) model to extract features from the images, and then applies a random forest classifier to predict the particle categories, providing a summary report of the particle classification results, including the overall accuracy and category-specific precision and recall values.
Bloodstain Pattern Analysis for Crime Scene Reconstruction
Using Google Maps and Earth Engine, analyze a set of bloodstain patterns at a crime scene to reconstruct the events surrounding the incident, by first georeferencing the bloodstain locations and then applying a spatial analysis algorithm to identify patterns and trajectories, considering factors such as stain size, shape, and distribution, and then using a physics-based model to simulate the bloodstain formation process, taking into account the angle of impact, velocity, and volume of blood, to estimate the location and movement of individuals at the scene, providing a 3D visualization of the reconstructed crime scene and a written report detailing the findings and limitations of the analysis.