ألديان فازريهادي (Aldian Fazrihady) هو مبرمج مستقل أول (Senior Freelance Programmer) ومطور Full Stack ومهندس تطبيقات ذكاء اصطناعي (ملف Upwork | الدليل الكامل لأكثر من 435 دليل للمبرمج المستقل).
AI platforms building autonomous data-science copilots, browser coding environments, or LLM code-interpreter agents that execute untrusted Python/JavaScript/Bash code safely frequently search to hire a freelance AI code execution sandbox (E2B, Modal, Firecracker MicroVMs, gVisor, Pyodide) programmer. I am Aldian Fazrihady, a Senior Freelance AI Application Engineer and Systems Security Architect available via Upwork.
Core Secure AI Code Interpreter & MicroVM Sandbox Capabilities
- Sub-150ms Ephemeral Sandboxes (E2B, Modal, Fly Machines, Firecracker): Spawning isolated microVM sandboxes pre-loaded with Pandas, Polars, Matplotlib, Playwright, and Scikit-learn where AI agents can run multi-step Python scripts, inspect stdout/stderr, and return generated charts or files safely.
- Strict Network Egress, CPU & Filesystem Isolation: Preventing container-escape, fork-bomb, and SSRF attacks via gVisor (`runsc`), seccomp-bpf profiles, cgroups v2 memory limits, and zero-trust egress allowlists.
- In-Browser Zero-Server Execution (Pyodide & WebContainers): Running Python scientific stacks directly in the user's browser WebAssembly sandbox when zero server compute cost is desired.
Frequently Asked Questions
Why is running LLM-generated Python inside standard `exec()` or a shared Docker container dangerous?
LLM-generated code (or prompt-injected user inputs) can read environment variables (`os.environ`), access internal VPC databases, or exhaust host RAM. Hardware-virtualized Firecracker microVMs or gVisor sandboxes isolate every session completely.
How can we hire you to architect our AI code-interpreter sandbox?
Message me on Upwork.