Building real-world AI solutions that cut through the noise (and reporting time)
I turn messy financial data into actionable insights using:
✅ Generative AI & ML (PyTorch/TensorFlow) → AI agents, math-driven models
✅ Python & AWS → Scalable pipelines, Power BI dashboards
✅ Trilingual precision → ES/EN/IT fluency for seamless collaboration
“Cut finance reporting time by 70%” at Comultrasan
→ Optimized ETL workflows + SQL data integrity
→ Democratized analytics for SMEs (Artyfal)
→ Military-grade reliability (10+ years engineering rigor)
| Category | Tools |
|---|---|
| AI/ML | PyTorch, TensorFlow, Scikit-Learn, H2O, PyCaret, BERT |
| Data | Python, SQL, Power BI, PySpark, pandas, AWS |
| Deployment | Gradio, Azure ML, Docker, CI/CD |
- Practical GenAI: Building explainable AI agents for finance/logistics
- Math-Driven Analytics: No hype — just linear regression, clustering, and anomaly detection that works
- Open Source: Check my repos for clean, documented code
I’m here for:
- Finance/Logistics Teams: Reduce reporting chaos with Power BI + Python
- Startups: MVPs that actually ship (AWS, Gradio, Streamlit)
- Ethical AI Projects: No black-hat tricks — just transparent, human-centered solutions
📩 Reach me: Upwork |
hedinyer.perucho [at] gmail.com
🌐 Timezone: Bogotá (GMT-5) — Flexible for EU/US hours
🎯 Pronouns: He/Him
⚡ Fun Fact: I speak 3 languages but still mix up “embarazada” and “imbarazzato” 😅
“Data isn’t about models — it’s about fixing what’s broken.”
— Veteran AI Engineer | Mechatronics B.Eng | AWS Certified