Roman Meclazcke presenting a product at Platanus

Hi, I'm

Roman
Meclazcke

Backend & AI Developer

Backend and AI developer with 3+ years building backends, RAG systems, and agents that reach production. I won the PlatanusHack public vote, maintain a Draw.io skill used by thousands of people, and share what I learn with a community of 16K+ developers.

Based in
Tandil, Argentina
Experience
3+ years in software

Experience

Work experience.

Professional background across companies and independent projects, with roles in backend development, automation, and applied AI development.

Index

TwinSF

AI Developer / Backend Developer

AI Developer

Present

Designed and implemented guardrails to evaluate production agents and reduce sensitive-data leakage. Built Mission Control, the agent operations and monitoring system used both internally and by clients. Also developed agents for support, sales, QA, and marketing, integrated with Slack, Google Calendar, and LinkedIn, and implemented RAG systems with LlamaIndex while evaluating LLM models per use case.

Backend Developer

Initial phase

Participated in the migration of legacy Oracle Forms applications to modern Spring Boot architectures, delivering a scalable transition for management systems used by public institutions. Worked across the full project cycle: requirements gathering, analysis, time estimation, and delivery. Designated as the responsible for onboarding new developers to the team.

Stack

  • Python
  • OpenClaw
  • Google Cloud
  • Spring Boot
  • Guardrails

Freelance

Backend Developer

  • Owned the full cycle: requirements gathering, analysis, design, development, and deployment of an automation system for a hair salon.
  • The solution let the client focus on core revenue-generating activities. Currently in production.

Stack

  • Node.js
  • TypeScript
  • PostgreSQL

Vopper

Backend Developer

  • Contributed to the backend development for the business logic of a logistics management platform. Worked with databases, performance optimization through caching, requirements analysis, and new features. Also worked with people from different areas to understand what they needed and implement improvements to the system.
  • The project centralized key processes for a supplier company: transport and route planning, per-store product dispatch, cash and account receivables management, and vehicle maintenance alerts. The solution helped organize daily operations, improve the follow-up of each process, and provide clearer information to respond to incidents.

Stack

  • NestJS
  • TypeScript
  • Redis
  • Docker
  • Kubernetes

UNICEN

University Technical Degree in Application Development

  • National University of the Center of Buenos Aires Province.

Projects

Selected work.

Products, tools, and technical builds where I worked on backend, AI, and real implementation.

FitVision

AI-powered virtual try-on platform for fashion ecommerce. Built, deployed, and operated independently.

Built solution

End-to-end deployed full-stack product with payments and AI processing in production.

AISpring BootReactPaymentsCloud

knowAhuman

Project built at PlatanusHack to clone your personality into an AI agent that can talk to other agents and detect valuable social relationships before human contact.

Built solution

Functional platform capable of cloning your personality into an AI agent and generating conversations with other agents.

HackathonAIAgentsOpenClawRailwaySupabase

Draw.io Skill for OpenClaw

OpenClaw skill that generates class, flow, sequence, and ERD diagrams using Draw.io.

Built solution

Published skill that generates Draw.io diagrams from structured instructions.

OpenClawDraw.ioDiagramsAutomationAI
Draw.io Skill for OpenClaw logo
3,448downloads

EatEligent

Healthy nutrition app built with NestJS, MySQL, Sequelize, and hexagonal architecture. Includes image uploads, AI processing, and email account activation.

Built solution

Backend with hexagonal architecture, email flows, image uploads, and AI-based filtering.

NestJSMySQLSequelizeEmail deliveryNSFW image filteringAI

RAG

RAG system designed to centralize internal documentation and enable fast queries over technical and operational knowledge.

Built solution

Searchable knowledge base using FastAPI, Qdrant, Postgres, and an assisted search interface.

PythonFastAPIQdrantPostgresAIStreamlit

Content Creator X

Codex skill that researches dev trends, evaluates topics with three subagents, and generates ready-to-post X content saved in Markdown.

Built solution

Skill with subagents, topic evaluation, and reproducible Markdown output.

CodexSubagentsAutomationGitHub TrendingMarkdownX

Stack

Technologies and tools I work with.

Tools I use to build production agents, backends, and systems that reach real users.

  • Java
  • Python
  • TypeScript
  • JavaScript
  • PHP
  • Node.js
  • Spring Boot
  • Spring Security
  • FastAPI
  • NestJS
  • Express
  • LangChain
  • LlamaIndex
  • OpenClaw
  • Hermes
  • MySQL
  • PostgreSQL
  • MongoDB
  • Oracle
  • Qdrant
  • Redis
  • Hibernate
  • Sequelize
  • Prisma
  • TypeORM
  • Google Cloud
  • Google Cloud VMs
  • AWS (EC2, S3, RDS, IAM)
  • Vercel
  • Docker
  • GitHub Actions
  • CI/CD
  • Hexagonal architecture
  • Event-driven architecture
  • Microservices
  • RAG
  • AI agents
  • Agent orchestration
  • Guardrails
  • Agent evaluation
  • Agent operations
  • Agent observability
  • External tool integration
  • Vector Search
  • Image processing
  • WebSockets
  • Kafka

Want to talk?

Leave me a message through the form and I'll reply by email. You can also find me on social media if you prefer.

Leave me a message

Complete the form and I'll reply by email.