AI Engineering Training: 1-on-1 Remote Sessions
Personalized training sessions via Google Meet. Master Claude Code, build AI agents with OpenClaw, learn advanced architecture patterns, fine-tune open models with LoRA/GRPO, and level up your DevOps skills. Adapted to your level -- no prior AI experience required.
By Jose Nobile | 19+ years experience | From $47.95/hr
Who Is This For?
Software Developers
Junior to senior developers who want to integrate AI into their daily workflow. Learn to use Claude Code effectively, write better CLAUDE.md files, and ship features 2x faster with AI assistance.
Tech Leads & Architects
Technical leaders who need to design AI-augmented workflows for their teams. Learn multi-agent architecture, MCP server development, and how to establish AI coding standards for your organization.
CTOs & Engineering Managers
Decision-makers evaluating AI tools for their engineering organization. Understand ROI, security implications (ZDR, SSO), cost optimization, and how to roll out Claude Code across teams of 10-100+ engineers.
Non-Coders Who Read Code
Product managers, designers, or entrepreneurs who can read code but don't write it daily. Learn to use AI assistants like OpenClaw for automation, and understand how AI agents can enhance your workflow.
Full 6-Block Curriculum
The curriculum is modular -- you can take any block independently or follow the full sequence. Each block consists of 2-4 sessions (2-8 hours), depending on depth and your pace. We work on your real projects, not toy examples. Block 6 (Fine-Tuning) is new for 2026 and covers the complete workflow from dataset preparation to production deployment.
AI Coding Fundamentals
Claude Code installation and configuration. Effective prompting techniques. Real project workflows. CLAUDE.md design for your codebase. Memory system and permission modes. Compact mode and cost management. By the end, you're using Claude Code productively on your own projects.
2-3 sessions | 4-6 hours
OpenClaw
Installation and architecture (gateway, nodes, channels). AGENTS.md, SOUL.md, and MEMORY.md design. WhatsApp and Telegram channel setup. Voice interaction configuration. Skills development. Cron jobs and scheduled tasks. Sub-agents. By the end, you have a personal AI assistant running on your preferred messaging platform.
3-4 sessions | 6-8 hours
Advanced Agents
Chain-of-thought reasoning and system prompt engineering. MCP server development (TypeScript and Python). Hook-driven automation. Multi-agent workflows and agent teams. Claude Agent SDK. Custom tool development with typed schemas. Structured outputs. By the end, you can build production-grade agent systems.
3-4 sessions | 6-8 hours
Software Engineering
SOLID principles with AI assistance. Clean Code practices. Test-Driven Development with AI (generate tests first, then implement). Unit, integration, and E2E testing strategies. Refactoring with AI. Code review automation. By the end, your code quality and test coverage are dramatically improved.
2-3 sessions | 4-6 hours
DevOps & Cloud
Docker containerization best practices. Kubernetes fundamentals (deployments, services, ingress). CI/CD pipeline design with GitLab CI or GitHub Actions. AWS or GCP basics. Helm chart management. Infrastructure as code. Monitoring and alerting. By the end, you can containerize and deploy applications with CI/CD.
2-3 sessions | 4-6 hours
Fine-Tuning & Model Customization
When prompting is not enough, fine-tune. Learn LoRA and QLoRA adapter training with Unsloth (2x faster, 80% less VRAM). Reinforcement learning with GRPO and preference alignment with DPO. Dataset preparation and curation. Free Google Colab workflows for Gemma 4, Llama, Qwen, and DeepSeek. Export to GGUF, vLLM, or Ollama. Deploy fine-tuned LoRA adapters on Cloudflare Workers AI or self-hosted vLLM. Fine-tuning vs RAG vs prompting: when to use each approach. By the end, you can customize any open model for your domain and deploy it to production.
3-4 sessions | 6-8 hours
April 2026 curriculum update: Block 3 now covers the Anthropic advisor pattern (pair Sonnet with Opus for 74.8% SWE-bench at lower cost than Opus solo) and Claude Managed Agents (fully hosted agent harness in public beta). Block 6 now includes Axolotl v0.29.0 with GRPO support -- the Tina paper showed LoRA fine-tuning with GRPO on just 7,000 examples can transfer DeepSeek R1's reasoning to smaller models for under $25 in compute. Fine-tuning a 7B model costs under $5 on modern cloud providers. All blocks reference Claude Opus 4.7 (released April 16, featuring xhigh effort level and task budgets for autonomous workloads).
How It Works
1. Free Consultation
We start with a free 1-hour Google Meet call. You tell me your goals, current level, and what you want to achieve. I assess where you are and propose a tailored learning path. No obligation.
2. Choose Your Plan
Pick a pricing tier that fits your budget. I recommend the 10-hour pack to start -- enough for 2 complete blocks. Sessions are scheduled flexibly around your availability.
3. Live Sessions
1-2 hour sessions via Google Meet with screen sharing. We work on your real projects -- not abstract exercises. I demonstrate techniques live, you practice immediately, and we solve your actual challenges together.
4. Async Support
Between sessions, you get async support via WhatsApp or Telegram for questions, code review, and troubleshooting. I provide curated resources, exercises, and reference materials tailored to your learning path.
Pricing
Payment: Binance (USDT) | PayPal | Credit/Debit card link | Prepaid packs
Frequently Asked Questions
Do I need to know how to code?
Not necessarily. You need to be able to read code (understand variables, functions, basic logic). The training is designed for people at all levels -- from non-coders who want to use AI assistants to senior developers building multi-agent systems.
What languages do you teach in?
Sessions are available in English or Spanish. Technical content uses industry-standard English terminology regardless of session language.
Can I take individual blocks instead of the full curriculum?
Absolutely. Each block is self-contained. Most people start with Block 1 (Claude Code) or Block 2 (OpenClaw) and continue based on their interests. The free consultation helps determine the best starting point.
What time zones do you work in?
I'm based in Colombia (UTC-5) and schedule sessions flexibly. I regularly work with clients in North America, Europe, and Latin America. We find times that work for both.
Are sessions recorded?
Yes, with your permission. All sessions are recorded via Google Meet and shared privately. You can re-watch techniques at your own pace and reference specific moments later.
What if I'm not satisfied after the first session?
The first hour is free -- no strings attached. For paid packs, if you're not satisfied after the first paid session, I'll refund the remaining unused hours. I'm confident the quality speaks for itself.
Do you offer team training?
Yes. For teams of 3+, I offer group sessions at a custom rate. Team training includes CLAUDE.md design for your specific codebase, workflow optimization, and hands-on workshops with your real projects.
Ready to Start?
Book your free 1-hour consultation. No commitment, no pressure. We'll discuss your goals, assess your current level, and I'll propose a personalized learning path. If it's a fit, we schedule your first session.