📫 How to reach me:
Software engineer with 10+ years of experience building and running products in production, from backend and infrastructure to frontend, with a recent focus on AI-powered systems. I own projects end to end: shaping the problem with product and design, delivering the code, and keeping it healthy afterwards. I enjoy mentoring: I have led development teams, run a community training course and taught at university.
Today I build LLM-based agents and document-review pipelines, and I run a fully AI-assisted development workflow (skills, git hooks, CI and automated review agents) that keeps quality high while automating everything that can be automated. Looking for greenfield, high-ownership projects where I can ship, teach and keep learning.
- Backend: Advanced Python and Node.js (Django, Django-Ninja, FastAPI, Flask, NestJS). Working knowledge of Go.
- Frontend: TypeScript with React and Angular for production web applications.
- Databases: Design, tuning and operation of SQL (PostgreSQL, MySQL, AWS RDS) and NoSQL databases.
- Cloud & DevOps: Docker, AWS (ECS, EC2, S3, Lambda, SQS/SNS), Terraform and AWS CDK, extensive CI/CD experience (GitHub Actions, Jenkins). Working experience with Kubernetes (EKS).
- Architecture & Distributed Systems: Monolithic and microservice architectures focused on scalability and performance; event-driven systems with RabbitMQ, SQS and Kinesis.
- AI & Machine Learning: LLM and agent systems (LangChain, LangGraph, Strands, Bedrock, Gemini, Langfuse, Ragas, MCP); classical ML (scikit-learn, XGBoost, Keras, PyTorch); MLOps (MLflow, Evidently); Streamlit for data apps.
- Engineering practices: AI-assisted SDLC with an agent harness and skills, test-driven quality gates, technical mentoring and team leadership.
ReluTech is a U.S.-based IT solutions provider specializing in reducing data center infrastructure costs for cloud-committed customers.
- AI-driven automation: Led and developed internal AI solutions to streamline processes and enhance service offerings. Researched and made the technical decisions in greenfield scenarios.
- Agentic systems: Built agents, chatbots, agentic workflows and data analytics to improve customer experience and operational efficiency.
- Cross-functional collaboration: Worked with several teams to drive the adoption of emerging technologies across the company.
- Ownership: Led, developed and maintained several company projects, such as RAMP and NerdRabbit.
- AI-assisted SDLC: Defined and run a development process where everything that can be automated is automated, and quality is enforced by tooling rather than by hand:
- Planning: skills that write PRDs, epics and tasks; Jira context pulled in through TWG, so the agent works from the real tickets.
- Development: skills that encode our conventions, so agents implement tasks consistently.
- Quality gates: git hooks that catch problems before they leave the machine, CI on every pull request, and tests and audits that verify each change.
- Review: automated review agents (Greptile and others) review every PR before a human does.
Projects
- ReluAdmin: Internal catalog of components and equipment; a full redesign and migration of the existing admin system.
- MiQ: New application workflow to manage internal quotes.
- Product Agent: ReAct agent that navigates, extracts and summarizes information to generate equipment descriptions automatically.
- EquipmentFinder: Detects duplicated equipment using classic NLP techniques and embeddings.
- Job Post Tracker: Crawls sites to extract job post titles and descriptions; trained a custom model to classify positions as IT or non-IT.
- Doc Split & Doc Review: LLM pipelines built with Gemini and LangChain. Doc Split divides a compliance document into checklists; Doc Review then reviews each checklist to audit whether the required information is present or missing.
Tech Stack: Python, Node, LangChain, LangGraph, Strands, Gemini, Bedrock, ChromaDB, FastMCP, Langfuse, Ragas, Evidently, MLflow, Docker, ECS, AWS CDK, GitHub Actions.
TrueML is a financial software company focused on improving customer experiences for distressed borrowers through innovative, data-driven solutions.
- Payments: Maintained and developed new features for the payment offer platform, integrating with USAePay. Tech Stack: Golang, Gin, AWS Lambda, DynamoDB, cdk8s.
- Compliance Checker: Contributed to a compliance framework that prevents non-compliant customer interactions. Tech Stack: Golang, Gorm, Gin, Kubernetes, EKS, Kinesis.
- Heartbeat: Contributed to the development and maintenance of a fully automated, reactive decision engine. Worked with the Data Science and Product teams to implement features that optimize payment offers using contextual bandits and data-driven heuristics. Tech Stack: Python, AWS (RDS, EC2, ECS, Kinesis), Jenkins.
Noq was an early-stage company developing IAMbic, an open-source platform for managing multi-cloud resources and permissions. IAMbic centralizes and simplifies cloud access, keeping a human-readable, bi-directional representation of identity and access management (IAM) in version control.
- IAMbic: Built features such as the AWS service control policies provider and integrated authentication with Okta. Stack: Python, Tornado, Docker, Git, Celery, AWS IAM & Organizations, Okta, SQS & SNS.
Selehann is a software development and technology consulting company. I collaborated with cross-functional teams to deliver software solutions.
- NerdRabbit: The first AI-powered, unbiased hiring platform, matching cloud-certified developers with long-term and hourly projects. As Software Engineer, I led the end-to-end development of the microservices application. Stack: Flask, MySQL, Redis, Docker, Angular, ElasticSearch, AWS (SQS, SNS, EC2, S3, ECS, Lex).
- Noble1: A company in the electronic and precision equipment repair and maintenance industry. As Software Engineer, I led the end-to-end development and migration of its new application. Stack: TypeScript, Angular, NestJS, MySQL, MongoDB, Redis, AWS (SQS, SNS, EC2, S3, ECS).
Seven years at Hexacta, advancing from technical roles to leading development teams and guiding key projects. My work covered Machine Learning, Data and Backend, always aimed at improving the quality and productivity of the company's projects. I also created and ran internal training courses and wrote technical posts for the Hexacta Engineering Medium blog.
- Ngage (1 year): Platform that gathers data from Zoom, Teams and Moodle to build real-time indicators for online classes. As Technical Leader, I led the team and handled its data-analytics and data-science problems: an ML model to infer indicators, complex formulas solved in near real time, and computer vision models that detect whether students are paying attention. Stack: C#, MediatR, Python, Azure ML, Docker, SignalR, React, RabbitMQ.
- HAT, Hexacta Architecture Team (2 years): Advised the technical managers of Hexacta's projects on architecture and design decisions, technologies and frameworks, to improve product quality and productivity. Assigned areas: Machine Learning, Data and Backend. Also wrote posts for the Hexacta Engineering Medium blog.
- Coalfire, Cybersecurity internal software (4 months): Software Engineer on an internal tool that runs background jobs for security tools to find vulnerabilities. Stack: Python, Django, Django REST Framework, PostgreSQL, Docker, Celery.
- Turner, Predictive Engine (10 months): Optimized channel broadcasting using hard rules and the predictions of an ML model. Led a team of two developers. Stack: Python, hypothesis testing, XGBoost, Spring Boot, Spring Data, Docker, Pandas.
- Hx Labs (1.5 years): Developed and ran an online training course for the community, with around 200 attendees per session. The course starts from a small architecture based on frontend/backend separation. I created the resources, the hands-on project and the presentations, and managed the course. Stack: .NET Core, React, Redux, LiteDB.
- Equinauts (2 months): Analytics solutions company. Identified and resolved bottlenecks and performance issues in PostgreSQL. Stack: Node.js, PostgreSQL.
- YPF, Extranets revamp (1.5 years): YPF is Argentina's state-owned oil and gas company. Led a small team of developers on the modernization of the system that manages lubricant orders for service station owners. Stack: C#, SQL, SOAP.
- Grupo Sancor Seguros, Suite migration (3 years): One of Argentina's leading insurance groups. Developer on several migration projects across core areas: medical exams, disease processes and claims, and monetary and legal aspects. Stack: C#, Knockout.js.
- Liberty Lion, Lead Engineer: Developed a fuel station management platform (HR, notifications). Stack: NestJS, Angular, PostgreSQL.
- Degustando, Founding Engineer (Backend & Machine Learning): Led backend and machine learning for a platform enabling delivery and takeaway services in Posadas City. Stack: MongoDB, Node.js, React, React Native, Python, Docker, AWS.
Assist in delivering lectures, guide students through practical exercises and explain data analysis, machine learning and visualization techniques. I design course materials, support hands-on labs and mentor students in the use of ML tools and methodologies. Stack: Kettle, Streamlit, Python, Machine Learning, ETL, databases and data warehousing.
Thesis in progress, on MLOps: a MovieLens recommender system with experiment tracking and model registry, drift and accuracy monitoring, and feedback-driven retraining with champion promotion and rollback. Stack: Python, PyTorch, MLflow, Evidently, FastAPI, PostgreSQL, Docker.
- Spanish — Native
- English — Intermediate (B2). Speak, read and write.