Bridging the gap between low-level physical hardware and high-concurrency cloud ecosystems—engineering deterministic embedded systems, robust RAG/agent architectures, and resilient backends.

Portfolio • LinkedIn • rodrygoleu7@gmail.com

• Software & Systems Engineer with a strong foundation in Clean Architecture, DDD, and event-driven backends (FastAPI, C# .NET, NestJS, Spring Boot).
• Applied AI & LLMs in Production: Architecting end-to-end RAG systems (FAISS, LangChain, Gemini API), on-device privacy-first models (Gemini Nano), and autonomous agent engines.
• Embedded & Edge Computing: Hands-on experience developing deterministic firmware (C++17, FreeRTOS concepts), cross-compiling for ARM (Raspberry Pi, ESP32-S3), and long-range RF telemetry (LoRa, XBee).
• Cloud & DevOps: Containerized microservices, multi-stage Docker builds, and automated CI/CD across GCP, Azure, and AWS.

An aerospace telemetry and payload transmission stack engineered under strict bandwidth and compute constraints.

• Bandwidth Optimization: Implemented a non-blocking fragmentation algorithm to interleave stereo JPEG fragments into a 5 Hz sensor telemetry stream over a narrow 433 MHz LoRa link.
• Deterministic Firmware (C++17): Designed a Finite State Machine (FSM) on a Raspberry Pi Zero 2W governed by sensor fusion thresholds (IMU + Altimeter) to isolate mission-critical states.
• Error Concealment: Ground station automatically reconstructs corrupted and dropped transmission packets using OpenCV inpainting, preserving aerial frames without dropping cycles.
• DevOps for Edge: Built a reproducible cross-compilation pipeline using Docker to target ARM architectures from x86 environments.

Production-grade RAG pipeline indexing the Peruvian legal framework with zero-hallucination constraints.

• Vector Knowledge Base: Chunked and indexed complete national statutes (Constitution, Civil Code) into FAISS with high-density semantic embeddings.
• Hallucination Mitigation: Implemented dynamic contextual grounding and statute injection, ensuring all LLM responses are cite-traceable to legal articles.
• Low-Latency Deployment: Packaged backend with multi-stage Docker builds and deployed to Google Cloud Run, drastically reducing image size and cold-start latency.

• Backend & APIs: Python (FastAPI, Flask), C# (.NET), TypeScript (NestJS, Node.js), Java (Spring Boot), REST, WebSockets
• Applied AI & ML: RAG, LangChain, Gemini API / Gemini Nano, FAISS, SciBERT, HDBSCAN, OpenCV, scikit-learn
• Embedded & Edge: C++17, ESP32 / ESP32-S3, Raspberry Pi, LoRa (SX1278), XBee, UART/Serial, FreeRTOS concepts
• Cloud & DevOps: Google Cloud (Cloud Run), Azure, AWS (S3), Docker, Terraform, GitHub Actions, Linux (Bash)
• Databases: PostgreSQL, PostGIS, MongoDB, SQLite, Vector Stores