Software & AI Engineer · Bachelor's in Computer Science (FIU) · CompTIA Security+ · Miami, FL
I build data-grounded systems that verify their own numbers. My focus is production AI — retrieval-grounded LLM applications where every figure reconciles to the source of truth — carried by an engineering discipline I bring to everything: clean module boundaries, deterministic pipelines, and tests that try to break the system before users do.
Most recently at Archangel, I shipped a governed AI analytics Copilot (LangChain · RAG · Node/React · PostgreSQL) that answers business questions only from curated data and writes prose, never numbers — so every reported figure reconciles exactly to the BI dashboards. Along the way I validated $99M+ in booked revenue by reconciling the new BI layer to the system of record, and presented the final build to the company CEO.
I work across the full abstraction stack, from GPU-level C++20 and OpenGL to enterprise CI/CD and cloud infrastructure.
In Silico Neural Dynamics — GPU-accelerated computational-neuroscience engine A 1,000-node Leaky Integrate-and-Fire network simulator built from scratch in C++20 and OpenGL. Physics decoupled from the render loop via a fixed-step accumulator, procedural graph topology via rejection sampling, O(N²) Euclidean synaptic wiring, and custom GLSL fragment shaders with SDF blending for membrane rendering.
ShieldClaw — evidence-backed vulnerability verification Validates Semgrep SQL-injection findings against owned Docker Compose apps inside hardened container execution, separating exploitable issues from theoretical ones — so security signal comes with proof, not guesses.
GPT Hardware Bridge — vision-driven autonomous robot An omnidirectional Raspberry Pi robot integrating GPT-4o Vision for object detection and speech recognition for voice control, on a hybrid Python/C++ architecture for optimized motor and camera control.
AI & ML: LLM applications, Retrieval-Augmented Generation, LangChain, LLM APIs (Claude, OpenAI, Ollama), prompt engineering, guardrails & evals Software & Full-Stack: Python, Java, C++20, C#, TypeScript, Node.js, React, REST APIs Data & BI: PostgreSQL, Oracle SQL, MS SQL, materialized views, ETL, data modeling, reconciliation Cloud & DevOps: AWS, Terraform, Docker, Kubernetes, CI/CD, GitHub Actions, Linux Security: CompTIA Security+, SAST (Semgrep), SQL-injection hardening, container sandboxing (seccomp), TDD Systems & Low-Level: OpenGL/GLSL, GPU pipeline engineering, manual memory management, CMake
- Recently earned my Bachelor's in Computer Science from Florida International University (Dean's List)
- Building a production AI analytics Copilot at Archangel Education + Technology
- Extending ShieldClaw with richer observer tiers and broader SAST support beyond Semgrep
- Adding inhibitory (GABAergic) neuron types to the neural simulator to stabilize network dynamics against hyper-synchronous firing
- Completing a Mobile Applications Development College Credit Certificate at Miami Dade College (Fall 2026)
bryanlondres5@gmail.com · LinkedIn · Miami, FL · open to new-grad software & AI roles



