N.º 01
Morris & Opazo — AWS Advanced Consulting Partner
AI Engineer Intern
- Piloting the team's first pilot of AWS's AI-DLC methodology with Spec-Driven Development (SDD) for a US fintech.
Enzo Loren.
Developer · Agentic · LangGraph · MCP · RAG · Evals
I specialize in agentic and multimodal systems, and in identifying where AI genuinely adds value: cutting costs, eliminating manual work, or making operationally unfeasible things viable.
§ 01
Experience
-
- N.º 02
Buses JAC
AI Engineer
- Designed and developed a GenBI Text-to-SQL assistant that allows managers to query business data in natural language, eliminating the wait for manual reports.
- Built the core AI pipeline using Vanna AI and DeepSeek V4, customizing the SQL execution layer to enforce business rules.
- Developed a containerized Flask web interface with built-in feedback loops to continuously train the model on undocumented edge cases.
- N.º 03
Morris & Opazo — AWS Advanced Consulting Partner
AI Engineer Intern
- Designed and deployed a multimodal agentic system on Strands Agents and AWS Bedrock AgentCore: speaker-diarized transcription (Silero VAD + ECAPA-TDNN), visual content extraction, and automated requirements evaluation. Processes 30-minute meetings in ~8.5 min at ~$0.34 USD.
- Built a custom transcription service achieving ~92% cost reduction vs. Amazon Transcribe, with API-compatible output.
- Developed a serverless MCP proxy (Lambda + API Gateway) for Asana integration, resolving an OAuth 3LO ↔ 2LO incompatibility.
- Conducted benchmarking of Amazon Nova 2 Omni across 5 modalities with 400+ tests under 4 extended reasoning levels.
- N.º 04
Morris & Opazo — AWS Advanced Consulting Partner
AI Engineer Intern
- Researched the state of the art in generative AI agents and prompt engineering techniques (structured XML, chain-of-thought, self-critique).
- Designed and implemented a multi-agent system using LangChain and AWS Bedrock that generates complete software specifications from natural language, featuring a reviewer agent, quality gates, and feedback loops.
§ 02
Technical stack
AI & ML
- LLM orchestration
- LangChain · Strands Agents · Claude Agents SDK
- Prompt engineering
- Structured XML · CoT · self-critique
- MCP protocol
- Serverless proxies · integrations
- RAG & vector databases
- Embeddings · retrieval
- Evals
- Benchmarks · quality gates
Cloud & Backend
- AWS
- Bedrock · AgentCore · Lambda · API Gateway
- CI/CD
- GitHub Actions
- Linux
- SSH · Bash
Languages & Tools
- Python
- Pipelines & agents
- SQL
- PostgreSQL
- Node.js · Git
- Tooling
Languages
- Spanish Native
- English C1 — EF SET 65/100
§ 03
Education & certifications
Education
-
Computer Engineering (Ingeniería Civil Informática)
Final yearUniversidad de La Frontera (UFRO)
Mar 2021 — Dec 2026
-
B.S. in Engineering Sciences
AwardedUniversidad de La Frontera (UFRO)
Mar 2021 — Jul 2025
Certifications
-
Agentic AI Advanced Learning Plan — Technical (Partner)
Amazon Web Services (AWS)
Issued Feb 2026 · Agentic AI development · Amazon Bedrock
-
Official EF SET English Certificate 65/100 (C1 Advanced)
EF SET
Issued Oct 2025 · Official certificate
§ 04 — Contact
Let's talk.
A project where AI must prove real value — not slides, results? Tell me what's operationally unfeasible today, and let's figure out how to make it viable.