Open to opportunities

MARLON DAVID BUSTOS PAUCAR

@marlondavidbustospau

Message

I build production-grade LLM/RAG systems and backend APIs that deliver reliable, deployable AI.

Peru

Message

What I'm looking for

I’m looking to build and deploy reliable LLM/RAG and agentic backend systems—partnering with teams that value traceability, robust REST APIs, and production engineering from cloud infrastructure to CI/CD.

I’m an AI Engineer and Software Engineer focused on production-grade LLM applications—especially LLM/RAG pipelines, agentic workflows, and the backend APIs that make them reliable in the real world. With 2+ years of experience, I integrate OpenAI-compatible and open-model stacks (OpenAI, Anthropic, Mistral, Hugging Face, Ollama, SGLang) into deployable systems.

At SENACE, I designed and deployed an end-to-end LLM pipeline for environmental admissibility evaluation, including OCR/VLM processing on GPU and a hybrid RAG approach (BM25 + dense embeddings with pgvector + Reciprocal Rank Fusion). I also built an LLM adjudication engine with structured prompting, few-shot examples, XML parsing, confidence scoring, and evidence citations to evaluate 76 regulatory criteria with traceability. Earlier, I delivered a production mHealth platform with offline-first mobile flows and secure backend validation, and I developed FootSelfie, a Flutter-based computer-vision product with cloud inference and dataset curation.

Experience

Work history, roles, and key accomplishments

Current

LLM/RAG Software Engineer

Current

SENACE – National Service for Environmental Certification

May 2026 - Present (1 month)

Designed and deployed an end-to-end LLM pipeline for environmental admissibility evaluation, processing large case files from MinIO/S3 with OCR/VLMs on GPU. Built a hybrid RAG system (BM25, pgvector, RRF) and an LLM adjudication engine to evaluate 76 regulatory criteria with structured outputs and evidence citations.

Hybrid RAG BM25 Dense Embeddings (Pgvector)Reciprocal Rank Fusion (RRF)fastAPI Fastify Structured Prompting & Evidence Citations RabbitMQ

mHealth Software Developer

IMPACT Salud

May 2025 - Jan 2026 (8 months)

Built a production-grade mHealth platform with an offline-first mobile app, backend APIs, and a web dashboard for clinical monitoring. Implemented a dynamic questionnaire engine with consistent validation across mobile and backend, plus secure offline multimedia upload and background synchronization via REST APIs.

React Native TypeScript Django REST Framework API Security PostgreSQL AWS Cloud Storage Workflows Clinical Questionnaires

Mobile App & AI Models Developer

Universidad Peruana Cayetano Heredia

Mar 2024 - Jan 2026 (1 year 10 months)

Developed FootSelfie, an end-to-end Flutter mobile app for detecting and monitoring conditions related to diabetic foot. Integrated computer vision with secure cloud inference, contributed to dataset curation/annotation, and improved stability with error handling, retry policies, and backend synchronization.

Flutter Dart Firebase Data Curation Camera Media Handling Offline Persistence & Retry Logic

Research Intern - Scheduling

University of Alberta

May 2025 - Aug 2025 (3 months)

Developed a workforce shift scheduling desktop application and optimization pipeline using Pyomo and the HiGHS solver. Modeled operational constraints and automated execution workflows with AWS and GitHub Actions while collaborating remotely with a Canada-based research team.

Workforce Optimization Pyomo HiGHS (Optimization Solver)GitHub Actions

IoT & AI Assistant

Innovation Laboratory – RUI

Jul 2024 - Mar 2025 (8 months)

Supported cross-functional initiatives across mobile development, IoT, and an AI assistant by contributing to coordination, planning, technical documentation, and prototyping. Helped execute development work through structured updates and blocker management.

Mobile Development AI Assistant Support Technical Documentation Cross functional Coordination Prototyping

IoT & Computer Vision Researcher

CTIC – UNI Smart Cities Lab

Mar 2022 - Jan 2023 (10 months)

Built computer vision prototypes for facial recognition and people tracking for smart-city-oriented systems. Conducted applied research and prototyping to support IoT and computer vision use cases.

Facial Recognition People Tracking Smart City Prototyping IoT Use Case Development