# Ale Zanello

> Agentic AI & GenAI Engineer based in Argentina. 6+ years shipping full-stack software, focused on production multi-agent systems with LangChain, LangGraph, and the Model Context Protocol (MCP). Author of Akashi, a multi-agent framework for Dart & Flutter. Available for freelance.

## About
- [Portfolio](https://azanello.com): Bento grid portfolio — projects, skills, certifications
- [Résumé](https://azanello.com/cv): Full CV — projects, experience, certifications
- [Recruiter brief](https://azanello.com/hire): One-page summary for recruiters — availability, location, key facts, experience, fit check
- [Résumé PDF](https://azanello.com/Alejandro_Zanello_Resume.pdf): Direct download, text-based PDF, regenerated hourly from live data
- [Résumé JSON](https://azanello.com/resume.json): Machine-readable résumé in the JSON Resume schema
- [Uses](https://azanello.com/uses): Tools and technologies for AI-powered development
- [Full long-form profile](https://azanello.com/llms-full.txt): Complete bio, projects, certifications, and contact info for LLM citation

## Experience
- Independent AI & Full-Stack Engineer — Freelance (06/2025 – Present): TypeScript, Python, Rust, Dart, LangGraph, Vertex AI, MCP, OAuth 2.1, Multi-Agent, Vercel AI SDK, PyTorch, Hugging Face, Reinforcement Learning, OpenTelemetry, Next.js, Supabase, Flutter, Firebase
- FlutterFlow Developer / AI Native Engineer — Somnio Software (10/2024 – Present): Python, TypeScript, Dart, LLMs, RAG, pgvector, Supabase, MCP, OAuth 2.1, Multi-Agent, Vercel AI SDK, Vertex AI
- Software Engineer — Detecta (07/2023 – 10/2024): Flutter, NestJS, Firebase, AWS
- LowCode Specialist — Detecta (08/2022 – 07/2023): FlutterFlow, Flutter, Firebase
- IT Manager — California Fitness (10/2017 – 11/2021): PHP, MySQL, Selenium, Cypress
- Branch Office Manager — California Fitness (01/2016 – 10/2017)

## Projects
- [Akashi](https://akashi.azanello.com): Provider-neutral agent framework for Dart & Flutter — a 14-package monorepo with 13 packages published on pub.dev, bringing the agent runtime primitives Python and TypeScript take for granted to the Dart ecosystem. Multi-agent orchestration via subagent-as-tool, control handoffs, and per-step model escalation.
- [NovaMX CRM](https://novamx.co): Real estate CRM with an LLM-powered lead/property classification engine. Public listings + internal lead tracker; agentic scoring runs on Edge Functions on top of Supabase.
- [Detecta Security Platform](https://detectasecurity.io): Triple-platform system for internal ops, customers, and drivers. GPS tracking via Wialon, Firebase BaaS, and AWS microservices.
- [mcp-deep-dive](https://github.com/AleSZanello/mcp-deep-dive): Protocol-first MCP reference in TypeScript: custom server + client + all 3 capabilities (tools, resources, prompts), both transports (stdio + HTTP), with wire-level JSON-RPC logging. No LangChain wrappers.
- [langgraph-langserve-starter](https://github.com/AleSZanello/langgraph-langserve-starter): Production-grade REST API template exposing a LangGraph agent via LangServe. Multi-node research agent with SqliteSaver thread-resume, streaming, Langfuse observability, zero-key fake-LLM mode.
- [local-rag-chatbot](https://github.com/AleSZanello/local-rag-chatbot): Fully-local RAG chatbot built with LangGraph + Chroma + Ollama. Multi-Query Fusion and RRF retrieval, SqliteSaver chat history, pluggable LLM provider. Zero API keys required.
- [langgraph-ts-agents](https://github.com/AleSZanello/langgraph-ts-agents): TypeScript port of LangGraph supervisor/worker patterns - streaming-first, fully typed with Annotation.Root + Zod, HITL approval, side-by-side Python references.
- [Somnio AI Tools](https://github.com/somnio-software/somnio-ai-tools): Open-source CLI that installs and runs AI agent skills, health checks, and security audits for Flutter and NestJS projects across Claude Code, Cursor, Gemini, and other AI coding CLIs.
- [Stores Helper](https://github.com/AleSZanello/stores-helper): MCP-style CLI exposing Play Store & App Store management (products, forms, subscriptions) as tools for AI agents. Lets coding agents read/write store data from the repo.
- [Spec Dev Ralph](https://github.com/AleSZanello/spec-dev-ralph): Starter template and guide for Spec Driven Development. Quick entry point for learning and applying the methodology.

## Blog
- [Shipping a Remote MCP Server That Works in Both ChatGPT and Claude: A Field Guide](https://azanello.com/blog/remote-mcp-server-chatgpt-claude-guide): How to ship a remote MCP server that ChatGPT and Claude both accept: tool design, JSON-RPC gotchas, OAuth 2.1, security and a two-host test matrix.
- [Who Signs Off When Agents Write the Code? Risk-Tiered Approval in an AI-First SDLC](https://azanello.com/blog/risk-tiered-approval-ai-first-sdlc): An AI-first SDLC approval workflow that grades each spec by risk, lets a reviewer agent sign low-risk work and sends sensitive changes to a human.
- [Layered AI Code Review: Commit-Time, PR-Time and Human Review That Don't Step on Each Other](https://azanello.com/blog/layered-ai-code-review-workflow): An AI code review workflow in three layers: a local reviewer on every commit, a PR reviewer that owns the verdict, and humans for judgement calls.
- [Bounded Autonomy: How to Let a Coding Agent Push Fixes Without Letting It Ship](https://azanello.com/blog/bounded-autonomy-coding-agents): AI coding agent guardrails that let an agent push review fixes on its own: armed autonomy, a cycle budget, a deterministic guard and a clean hand-back.
- [Separate Talking From Acting: A Guardrail Architecture for Production AI Agents](https://azanello.com/blog/ai-agent-guardrails-architecture): AI agent guardrails that hold in production: stage every turn, make tools the only write path and guard what the agent promises before it's sent.
- [Your CLAUDE.md Is Probably a Changelog: Agent Instruction Files That Scale](https://azanello.com/blog/claude-md-agent-instruction-files-best-practices): CLAUDE.md best practices for long-lived repos: audit your agent instruction file, move history out, split decisions into ADRs, and generate status.
- [LLMs at the Edges: Designing Conversational Agents With a Deterministic Core](https://azanello.com/blog/conversational-agent-deterministic-flow-llm): How to build a deterministic conversational agent with LLM extraction at the edges: a pure flow runner, cheap models for parsing and human takeover.
- [ADRs and Session Handoffs: Giving AI Coding Agents a Memory That Survives](https://azanello.com/blog/adrs-session-handoffs-ai-coding-agents): Architecture decision records for AI coding agents, plus a session handoff file, let a fresh session pick up fast and stop re-arguing settled choices.
- [Evaluate What Your Agent Does, Not What It Says: A Guide to Tool-Call Evals](https://azanello.com/blog/agent-tool-call-evals-guide): AI agent evals that check tool calls, not prose: expected tools, forbidden tools, grounded facts, and guardrails that prune claims no tool backed.
- [RAG Over Your Own Docs With pgvector: Chunking, Retrieval and Answers You Can Trust](https://azanello.com/blog/rag-over-your-docs-with-pgvector): A practical RAG pgvector guide for your own docs: chunking, hybrid search, citations, index freshness and evals that teach the bot to say I don't know.
- [Human Handoff for AI Chat Agents: A Practical Design Guide](https://azanello.com/blog/ai-agent-human-handoff-guide): AI agent human handoff, done right: what a chat agent may do alone, what it only drafts, and how to pass a conversation to a person without losing context.
- [Red-Teaming Your LLM Agent: Prompt Injection and Prompt Leakage Tests](https://azanello.com/blog/prompt-injection-testing-llm-agents): A practical guide to prompt injection testing for LLM agents: leakage probes, indirect injection via RAG, image exfiltration and a CI gate.
- [How to Build an Eval Gym for Conversational AI Agents](https://azanello.com/blog/eval-gym-for-conversational-ai-agents): A practical guide to AI agent evaluation: four eval layers, LLM-as-judge scoring, N-run noise checks and simulated customers for chat agents.
- [A Week of Spec-Driven Development With GSD](https://azanello.com/blog/gsd-workflow-real-project): An honest week-long diary of shipping features on a real project with GSD. Phases, commits, time spent, and what broke along the way.
- [Migrating From OpenSpec to GSD: When It's Worth It](https://azanello.com/blog/migrating-from-openspec-to-gsd): Thinking about switching from OpenSpec to GSD? Here's an honest look at what you gain, what you give up, and when the migration is worth the cost.
- [What Is Spec-Driven Development? A 2026 Primer](https://azanello.com/blog/what-is-spec-driven-development): Spec-driven development (SDD) is a methodology where specs come before code. Here's what it is, why AI agents made it practical, and how to start.
- [GitHub Spec Kit: A Hands-On Review for 2026](https://azanello.com/blog/github-spec-kit-review): GitHub Spec Kit review after shipping real features with it. The Constitution, the 7 phases, the ceremony tax, and when it's worth the cost.
- [OpenSpec Explained: A Hands-On Guide for 2026](https://azanello.com/blog/openspec-deep-dive): OpenSpec is a lightweight spec-driven development framework for brownfield codebases. How it works, when to use it, and where it falls short.
- [Inside GSD 2.0: The Architecture Behind Reliable AI Agents](https://azanello.com/blog/gsd-2-architecture-deep-dive): A technical deep dive into GSD 2.0's architecture: anchor pruning, deterministic tools, fractal summaries, and the engineering that makes AI agents reliable.
- [Spec-Driven Development: GSD vs Spec Kit vs OpenSpec](https://azanello.com/blog/spec-driven-development-tools-compared): A practitioner's comparison of the three leading spec-driven development frameworks for AI coding agents. Architecture, trade-offs, and honest recommendations.
- [Claude Code Agents and Skills: A Practical Guide](https://azanello.com/blog/claude-code-agents-skills-guide): Learn how Claude Code agents and skills work, how to create custom workflows, and best practices for getting the most out of AI-assisted development.
- [Spec-Driven Development with GSD: A Complete Guide](https://azanello.com/blog/spec-driven-development): Spec-Driven Development turns AI coding from chaotic prompting into systematic engineering. Learn how GSD works step by step with real examples.

## Certifications (3 credentials, 19 course certificates — all verifiable)
- [IBM Generative AI Engineering Professional Certificate](https://www.coursera.org/account/accomplishments/specialization/DT7I0XVP5JQZ): IBM via Coursera, October 2026, 16 courses
  - [AI Agents Using RAG and LangChain](https://www.credly.com/badges/97306a60-b4c6-4fa0-a61f-6fe39a32e176): October 2026
  - [Generative AI Applications Specialist](https://www.credly.com/badges/f6ca33c9-0110-4e74-89a6-c36728d445bc): October 2026
  - [Generative AI Advanced Fine-Tuning for LLMs](https://www.credly.com/badges/b0c909a3-9559-477f-9bc2-50a4fa40145a): October 2026
  - [Generative AI Engineering with Transformers & LLMs](https://www.credly.com/badges/96badce6-3d3c-45ea-b162-d5c1092e030b): October 2026
  - [Deep Learning Essentials with Keras](https://www.credly.com/badges/138f7811-c12d-4221-a435-2c3cd2ffaf94): October 2026
  - [Generative AI and LLMs: Architecture and Data Preparation](https://www.credly.com/badges/f8191566-1df1-4c39-a4fb-40a1e7453399): October 2026
  - [Generative AI Foundational Models for NLP & Language Understanding](https://www.credly.com/badges/d9134bd0-ed3b-49ff-a2a0-922f955ccac4): October 2026
  - [Generative AI Language Modeling with Transformers](https://www.credly.com/badges/d100c180-6965-44c9-bcb9-d38e34708313): October 2026
  - [Data Analysis with Python](https://www.credly.com/badges/aeba365e-8f08-4a16-b19c-e1bb684b83f6): October 2026
  - [Machine Learning with Python (V2)](https://www.credly.com/badges/955ce890-b565-4943-82c3-1c7ea8531154): October 2026
  - [Python Project for AI and Application Development](https://www.credly.com/badges/3e65ebc3-41b1-4d7c-92fd-43a73e6089a7): June 2026
  - [Building Generative AI-Powered Applications with Python](https://www.credly.com/badges/9a382908-73ab-4584-a26a-648f2f8d8ff1): June 2026
  - [Artificial Intelligence Essentials V2](https://www.credly.com/badges/a05a8a27-a38b-4988-bc00-4255206357f3): June 2026
  - [Generative AI Essentials](https://www.credly.com/badges/333b935e-5eb8-4708-8b25-293d69d1594a): June 2026
  - [Generative AI: Prompt Engineering](https://www.credly.com/badges/dc139939-0e30-4c27-8d47-299a7b2154d4): June 2026
- [Agentic AI Engineering Specialization](https://www.coursera.org/account/accomplishments/specialization/V20XQMM270EL): Edureka via Coursera, May 2026, 4 courses
  - [Developing MCP-Powered Agentic AI Systems](https://www.coursera.org/account/accomplishments/verify/AXFAZWOU3U40): May 2026
  - [Multi-Agent Systems with LangGraph](https://www.coursera.org/account/accomplishments/verify/POMUZ7UGQNBS): May 2026
  - [Applied Agentic AI Pipelines with LangChain](https://www.coursera.org/account/accomplishments/verify/10KPCYZ9UVW9): May 2026
  - [Building Your First AI Agent with LangChain](https://www.coursera.org/account/accomplishments/verify/12NAU4R4NZEB): May 2026
- [AWS Certified Cloud Practitioner](https://www.credly.com/badges/a1c0c051-7d10-4633-86bb-7d9949b01615): Amazon Web Services, May 2024
- [Credly profile](https://www.credly.com/users/azanello): All badges

## Knowledge graph
- [Interactive graph](https://azanello.com/graph): every role, project, skill, credential and post as a linked note

### Person
- [Ale Zanello](https://azanello.com/graph/ale-zanello): I'm an AI Engineer in Córdoba, Argentina — agent runtimes, multi-agent systems and GenAI, with 6+ years shipping mobile, backend and web software.

### Role
- [California Fitness](https://azanello.com/graph/california-fitness): IT Manager (10/2017 – 11/2021) and Branch Office Manager (01/2016 – 10/2017) at California Fitness, Cancún — custom PHP/MySQL CRM and QA automation.
- [Detecta](https://azanello.com/graph/detecta): Software Engineer (07/2023 – 10/2024) and LowCode Specialist (08/2022 – 07/2023) at Detecta, Mexico (remote) — built the Detecta Security Platform.
- [Independent AI & Full-Stack Engineer (Freelance)](https://azanello.com/graph/freelance): Freelance since 06/2025, remote — conversational AI agents, real-time data systems, white-label Flutter apps and Next.js + Supabase products.
- [Somnio Software](https://azanello.com/graph/somnio-software): FlutterFlow Developer / AI Native Engineer at Somnio Software since 10/2024, remote — internal tooling and platform work.

### Line of work
- [AI Agent Products & Remote MCP](https://azanello.com/graph/agent-products-remote-mcp): Production AI agents on Vertex AI and LangGraph, exposed through a remote MCP server with its own OAuth 2.1 authorization server.
- [Conversational AI Agents](https://azanello.com/graph/conversational-ai-agents): Multi-tenant conversational AI agents in a TypeScript monorepo — Hono gateway, LangGraph.js with Postgres checkpointing, MCP, tracing and evals.
- [Real-time Data Systems](https://azanello.com/graph/realtime-data-systems): Event-driven data systems in Python and Rust, with ML research on top — time-series foundation models, PPO and evolution strategies.
- [SaaS Products for Small Businesses](https://azanello.com/graph/smb-saas-products): Next.js + Supabase products for small businesses — row-level security, Edge Functions, MercadoPago payments with webhooks, PWAs, reporting and ETL.
- [White-label Flutter Apps](https://azanello.com/graph/white-label-flutter-apps): 6+ white-label Flutter apps shipped from shared codebases — multi-flavor builds, Bloc/Riverpod, RevenueCat, Firebase/Supabase, Codemagic and fastlane CI/CD.

### Project
- [Akashi](https://azanello.com/graph/akashi): Provider-neutral agent framework for Dart and Flutter — multi-agent orchestration, durable execution and human-in-the-loop, published on pub.dev.
- [Detecta Security Platform](https://azanello.com/graph/detecta-security-platform): Triple-platform system for internal operations, customers and field drivers at Detecta — Wialon GPS tracking, Firebase and NestJS microservices on AWS.
- [FlutterFlow issue triage](https://azanello.com/graph/flutterflow-issue-triage): Core contributor to FlutterFlow's public GitHub issue triage at Somnio Software, covering 600+ community bug reports.
- [langgraph-langserve-starter](https://azanello.com/graph/langgraph-langserve-starter): REST API template exposing a LangGraph research agent via LangServe — SQLite thread resume, streaming, Langfuse tracing and a zero-key fake-LLM mode.
- [langgraph-ts-agents](https://azanello.com/graph/langgraph-ts-agents): Strictly typed TypeScript port of LangGraph supervisor/worker patterns — streaming-first, Zod-validated, with a human-in-the-loop approval gate.
- [local-rag-chatbot](https://azanello.com/graph/local-rag-chatbot): Fully local RAG chatbot on LangGraph, Chroma and Ollama — Multi-Query Fusion with Reciprocal Rank Fusion, SQLite chat history, zero API keys.
- [mcp-deep-dive](https://azanello.com/graph/mcp-deep-dive): Protocol-first MCP reference in TypeScript — a custom server and client, all three capabilities, stdio and HTTP transports, wire-level JSON-RPC logging.
- [NovaMX CRM](https://azanello.com/graph/novamx-crm): Real estate CRM for NovaMX — public property listings plus an internal lead tracker, with an LLM-powered lead and property classification engine.
- [Somnio AI Tools](https://azanello.com/graph/somnio-ai-tools): Open-source CLI that installs and runs AI agent skills, health checks and security audits across AI coding CLIs. I'm the
- [Spec Dev Ralph](https://azanello.com/graph/spec-dev-ralph): Ralph Loop — a starter template for spec-driven development where four Claude Code agents write specs, build, test and review in an autonomous loop.
- [Stores Helper](https://azanello.com/graph/stores-helper): Agent-friendly CLI scripts for App Store Connect, Google Play and Firebase — products, subscriptions and metadata managed from the repo by AI coding agents.

### Skill
- [Anthropic Claude](https://azanello.com/graph/anthropic-claude): Anthropic's Claude models and SDK — integrated in Akashi, my MCP work and labs, and my daily coding agent via Claude Code.
- [Apache Parquet & Arrow](https://azanello.com/graph/parquet): Columnar file and memory formats — the storage layer for my event histories, read with DuckDB and PyArrow.
- [AWS](https://azanello.com/graph/aws): Amazon Web Services — NestJS microservices at Detecta, self-terminating EC2 Spot research jobs, and the AWS Cloud Practitioner credential.
- [Bloc](https://azanello.com/graph/bloc): Event-driven state management pattern for Flutter — part of the architecture of my white-label apps.
- [C](https://azanello.com/graph/c-lang): Low-level systems language — I use C microbenchmarks to measure fan-out strategies before choosing a design.
- [Claude Code](https://azanello.com/graph/claude-code): Anthropic's agentic coding CLI — my daily driver for spec-driven development, with custom agents and skills, and a topic I write about.
- [Cloudflare Workers](https://azanello.com/graph/cloudflare-workers): Edge serverless functions — I used Workers to probe network latency from different edge locations.
- [Codemagic](https://azanello.com/graph/codemagic): CI/CD for mobile apps — I use it to build and ship multi-flavor Flutter apps.
- [Cursor](https://azanello.com/graph/cursor): AI-first code editor and CLI — part of my AI coding toolkit, supported by tooling I've contributed to.
- [Cypress](https://azanello.com/graph/cypress): JavaScript end-to-end testing framework — part of the QA automation I introduced at California Fitness.
- [Dart](https://azanello.com/graph/dart): The language behind Flutter — I authored Akashi, a 14-package Dart agent framework (13 on pub.dev), and ships white-label Flutter apps.
- [Docker](https://azanello.com/graph/docker): Containers for packaging and deploying services — used for my real-time data system deployments.
- [DuckDB](https://azanello.com/graph/duckdb): Embedded analytical SQL engine — I use it over Parquet to query large event histories without running a database server.
- [Evals](https://azanello.com/graph/evals): Systematic evaluation of LLM and agent behaviour — test sets, graders and regression checks that make non-deterministic systems shippable.
- [FastAPI](https://azanello.com/graph/fastapi): Async Python web framework — I use it to serve LangGraph agents and to build real-time, event-driven backends.
- [fastlane](https://azanello.com/graph/fastlane): Automation for iOS and Android builds, signing and store releases — part of my white-label app pipeline.
- [Firebase](https://azanello.com/graph/firebase): Google's app backend platform — the BaaS behind the Detecta Security Platform and several of my white-label apps.
- [Flutter](https://azanello.com/graph/flutter): Google's cross-platform UI toolkit — I ship white-label Flutter apps and built Flutter-native agent tooling in Akashi.
- [FlutterFlow](https://azanello.com/graph/flutterflow): Visual low-code builder that generates Flutter code — I shipped production apps with it and triaged 600+ public FlutterFlow issues.
- [GCP](https://azanello.com/graph/gcp): Google Cloud Platform — part of my cloud toolkit, used to provision Firebase projects for multi-flavor apps.
- [GitHub Actions](https://azanello.com/graph/github-actions): GitHub's CI/CD and automation platform — runs the multi-job test pipeline of my Akashi monorepo.
- [Google Colab](https://azanello.com/graph/google-colab): Hosted Jupyter with GPUs — where I run fine-tuning jobs, with datasets pulled from Hugging Face.
- [Google Gemini](https://azanello.com/graph/google-gemini): Google's Gemini models (and on-device Gemma) — the default provider in several of my open-source agent projects.
- [Google Vertex AI](https://azanello.com/graph/vertex-ai): Google Cloud's managed AI platform — I run Gemini-powered LangGraph agents on Vertex AI in production.
- [GraphQL](https://azanello.com/graph/graphql): Query language for APIs — part of my backend toolkit; this site uses GitHub's GraphQL API for contribution data.
- [Hono](https://azanello.com/graph/hono): Lightweight, standards-based TypeScript web framework — the gateway layer of my multi-tenant conversational agents.
- [Hugging Face](https://azanello.com/graph/hugging-face): The open model hub and Transformers ecosystem — I version my ML datasets there, and studied its libraries in depth in the IBM certificate.
- [LangChain](https://azanello.com/graph/langchain): Framework of composable LLM building blocks — models, prompts, tools and retrievers — that I use under my LangGraph labs.
- [Langfuse](https://azanello.com/graph/langfuse): Open-source LLM observability platform — traces, costs and evaluations; I wire it into LangGraph services.
- [LangGraph](https://azanello.com/graph/langgraph): Stateful graph orchestration for LLM agents — my default for multi-step and multi-agent workflows in Python and TypeScript.
- [LangSmith](https://azanello.com/graph/langsmith): LangChain's platform for tracing, debugging and evaluating LLM applications — part of my observability and evals toolkit.
- [Linux & systemd](https://azanello.com/graph/linux-systemd): Running long-lived services on a Linux VPS — I operate daemons with systemd units, restart policies and scheduled maintenance.
- [LLMs](https://azanello.com/graph/llms): Large language models — the core building block behind every agent, RAG pipeline and AI product I ship.
- [Maestro](https://azanello.com/graph/maestro): Declarative end-to-end UI testing for mobile apps — part of my mobile testing toolkit.
- [Make](https://azanello.com/graph/make): Visual automation platform (formerly Integromat) — part of my low-code and automation toolkit.
- [MCP (Model Context Protocol)](https://azanello.com/graph/mcp): Open JSON-RPC protocol that connects AI agents to tools, resources and prompts — I have built servers, clients and adapters for it.
- [MercadoPago](https://azanello.com/graph/mercadopago): Latin America's leading payments platform — integrated with webhooks in my Next.js + Supabase SMB products.
- [Multi-Agent Systems](https://azanello.com/graph/multi-agent): Designing systems where several LLM agents delegate, hand off and cooperate — a core focus of my open-source work.
- [MySQL](https://azanello.com/graph/mysql): Open-source relational database — the backend of the custom PHP CRMs I built and maintained.
- [n8n](https://azanello.com/graph/n8n): Open-source workflow automation tool — part of my low-code and automation toolkit.
- [NestJS](https://azanello.com/graph/nestjs): Structured Node.js backend framework — I built AWS microservices with it for the Detecta Security Platform.
- [Next.js](https://azanello.com/graph/nextjs): React framework for full-stack web apps — my choice for SMB SaaS products, NovaMX CRM and this site.
- [Node.js](https://azanello.com/graph/nodejs): JavaScript runtime for servers and tooling — under my NestJS microservices, Hono gateway and TypeScript agent work.
- [OpenAI](https://azanello.com/graph/openai): OpenAI's GPT models and API — one of the providers I integrate, from NovaMX's classification engine to Akashi's OpenAI adapter.
- [OpenTelemetry](https://azanello.com/graph/opentelemetry): Vendor-neutral tracing standard — I ship an OpenTelemetry exporter in Akashi and trace agents with it in client work.
- [pgvector](https://azanello.com/graph/pgvector): PostgreSQL extension for storing embeddings and running vector similarity search — the vector store behind RAG on Postgres and Supabase.
- [PHP](https://azanello.com/graph/php): Server-side web language — I built and maintained a fully custom PHP/MySQL CRM for several years.
- [Playwright](https://azanello.com/graph/playwright): Cross-browser end-to-end testing and automation framework — part of my testing toolkit.
- [PostgreSQL](https://azanello.com/graph/postgresql): The relational database under my Supabase products, agent checkpointing and vector search.
- [Prisma](https://azanello.com/graph/prisma): Type-safe ORM for Node.js and TypeScript — part of my backend toolkit.
- [Python](https://azanello.com/graph/python): The language of the AI ecosystem — I use it for LangGraph services, RAG pipelines, real-time backends, ML research and automation scripts.
- [PyTorch](https://azanello.com/graph/pytorch): Deep learning framework I use for real model work — fine-tuning a time-series foundation model and PPO implemented from scratch.
- [RAG (Retrieval-Augmented Generation)](https://azanello.com/graph/rag): Grounding LLM answers in retrieved documents — I've built RAG pipelines in Python and a provider-neutral RAG package in Dart.
- [React](https://azanello.com/graph/react): Component-based UI library — the foundation of my Next.js products and this site.
- [RevenueCat](https://azanello.com/graph/revenuecat): In-app subscription infrastructure for iOS and Android — wired into my white-label Flutter apps.
- [Riverpod](https://azanello.com/graph/riverpod): Compile-safe state management and dependency injection for Flutter — used in my white-label app codebases.
- [Rust](https://azanello.com/graph/rust): Systems language for safe, fast code — I use it for the WebSocket feed and message bus on the low-latency path of a real-time data system.
- [scikit-learn](https://azanello.com/graph/scikit-learn): Classical machine learning in Python — I use it and gradient-boosted trees as baselines that foundation models have to beat.
- [Selenium](https://azanello.com/graph/selenium): Browser automation framework — I introduced Selenium-based QA automation for a custom CRM and mobile apps.
- [Supabase](https://azanello.com/graph/supabase): Open-source Postgres backend platform — auth, row-level security, Edge Functions and pgvector; my default BaaS for web products.
- [Svelte](https://azanello.com/graph/svelte): Compiled UI framework — I build lightweight real-time monitoring dashboards with Svelte and Vite.
- [Swift](https://azanello.com/graph/swift): Apple's language — I built a small macOS companion app that surfaces a running system's live state.
- [Tailwind CSS](https://azanello.com/graph/tailwind-css): Utility-first CSS framework — part of my front-end stack, used to style this site.
- [TypeScript](https://azanello.com/graph/typescript): Typed JavaScript — my main language for MCP, LangGraph.js agents, Next.js products and web tooling.
- [Vercel](https://azanello.com/graph/vercel): Frontend cloud for Next.js — hosts this site and maintains the Vercel AI SDK in my TypeScript stack.
- [Vercel AI SDK](https://azanello.com/graph/vercel-ai-sdk): TypeScript toolkit for streaming LLM apps and tool-calling agents — part of my TypeScript AI stack and an influence on Akashi's API.

### Concept
- [Agent Orchestration](https://azanello.com/graph/agent-orchestration): Coordinating LLM agents, tools and steps into reliable workflows — the hub concept behind most of my AI engineering.
- [Deep Learning](https://azanello.com/graph/deep-learning): Machine learning with multi-layer neural networks — the foundation under LLMs, from my IBM coursework to PyTorch models I train myself.
- [Durable Execution](https://azanello.com/graph/durable-execution): Persisting agent state at every step so runs survive crashes, pause for approval and resume later — a core pillar of Akashi.
- [Education & Languages](https://azanello.com/graph/education): Systems Engineering, a Mobile App Development diploma and a Bachelor's in International Business, all in Mexico; Spanish native, English full professional.
- [Evolution Strategies](https://azanello.com/graph/evolution-strategies): Gradient-free policy optimization — I used OpenAI-style ES with antithetic sampling when PPO could not assign credit through noisy rewards.
- [Fine-Tuning](https://azanello.com/graph/fine-tuning): Adapting a pretrained model with further training — I fine-tuned a time-series foundation model on GPU; LLM methods like LoRA and DPO come from coursework.
- [Generative UI](https://azanello.com/graph/generative-ui): Letting the model render interface, not just text — I streamed React components from tool calls with the Vercel AI SDK.
- [Human-in-the-Loop](https://azanello.com/graph/human-in-the-loop): Pausing an agent for a person to approve, edit or reject an action before it continues — built into Akashi and my LangGraph labs.
- [NLP](https://azanello.com/graph/nlp): Natural language processing — from n-grams and embeddings to sequence-to-sequence and transformer models, studied in my IBM coursework.
- [OAuth 2.1](https://azanello.com/graph/oauth-2-1): Modern authorization for apps and AI clients — I built an OAuth 2.1 authorization server for a remote MCP server: DCR, PKCE, rotation.
- [Observability](https://azanello.com/graph/observability): Tracing, token accounting and evals that make LLM systems debuggable in production — built into my agent work.
- [Prompt Engineering](https://azanello.com/graph/prompt-engineering): Designing instructions, examples and context so LLMs behave reliably — formally studied and applied across my agents and specs.
- [Reinforcement Learning](https://azanello.com/graph/reinforcement-learning): Learning decision policies from reward — I implemented PPO from scratch in PyTorch against custom simulation environments.
- [Spec-Driven Development](https://azanello.com/graph/spec-driven-development): Writing a structured spec before any code so AI coding agents execute against it — the methodology I practise, template and write about.
- [Time-Series Foundation Models](https://azanello.com/graph/time-series-foundation-models): Pretrained models for sequential numeric data — I fine-tuned IBM Granite TSPulse on GPU and benchmarked it against tabular baselines.
- [Tool Calling](https://azanello.com/graph/tool-calling): Letting an LLM call typed functions and APIs in a loop — the mechanism that turns a chat model into an agent.
- [Transformers](https://azanello.com/graph/transformers): The attention-based neural architecture behind modern LLMs — studied hands-on through the IBM Generative AI Engineering certificate.
- [White-Labeling](https://azanello.com/graph/white-labeling): Shipping many branded apps from one shared codebase — flavors, per-brand config and automated store pipelines.

### Credential
- [Agentic AI Engineering Specialization](https://azanello.com/graph/agentic-ai-engineering): Edureka's four-course Coursera specialization on building agents with LangChain, LangGraph and MCP. Completed May 2026.
- [AWS Certified Cloud Practitioner](https://azanello.com/graph/aws-cloud-practitioner): Amazon Web Services' foundational cloud certification — core services, security, pricing and architecture. Earned May 2024.
- [IBM Generative AI Engineering Professional Certificate](https://azanello.com/graph/ibm-generative-ai-engineering): IBM's 16-course professional certificate on Coursera — LLMs, transformers, fine-tuning, RAG and LangChain. Completed October 2026.

### Course
- [AI Agents Using RAG and LangChain](https://azanello.com/graph/ai-agents-using-rag-and-langchain): Builds LLM applications grounded in retrieved documents. IBM via Coursera, October 2026.
- [Applied Agentic AI Pipelines with LangChain](https://azanello.com/graph/applied-agentic-ai-pipelines-with-langchain): ReAct agents, memory and retrieval pipelines with LangChain. Edureka via Coursera, May 2026.
- [Artificial Intelligence Essentials V2](https://azanello.com/graph/artificial-intelligence-essentials-v2): AI concepts, business use, careers and ethics. IBM via Coursera, June 2026.
- [Building Generative AI-Powered Applications with Python](https://azanello.com/graph/building-generative-ai-powered-applications-with-python): Generative AI applications and chatbots in Python. IBM via Coursera, June 2026.
- [Building Your First AI Agent with LangChain](https://azanello.com/graph/building-your-first-ai-agent-with-langchain): A first tool-using, retrieval-backed agent with LangChain. Edureka via Coursera, May 2026.
- [Data Analysis with Python](https://azanello.com/graph/data-analysis-with-python): Data wrangling and analysis with NumPy, Pandas and SciPy. IBM via Coursera, October 2026.
- [Deep Learning Essentials with Keras](https://azanello.com/graph/deep-learning-essentials-with-keras): Neural network fundamentals with Keras. IBM via Coursera, October 2026.
- [Developing MCP-Powered Agentic AI Systems](https://azanello.com/graph/developing-mcp-powered-agentic-ai-systems): Building agentic systems on the Model Context Protocol. Edureka via Coursera, May 2026.
- [Generative AI Advanced Fine-Tuning for LLMs](https://azanello.com/graph/generative-ai-advanced-fine-tuning-for-llms): Instruction tuning and preference optimization (DPO) for LLMs. IBM via Coursera, October 2026.
- [Generative AI and LLMs: Architecture and Data Preparation](https://azanello.com/graph/generative-ai-and-llms-architecture-and-data-preparation): LLM architectures and text data preparation with PyTorch and Hugging Face. IBM via Coursera, October 2026.
- [Generative AI Applications Specialist](https://azanello.com/graph/generative-ai-applications-specialist): Builds generative AI applications with LangChain, embeddings and Gradio. IBM via Coursera, October 2026.
- [Generative AI Engineering with Transformers & LLMs](https://azanello.com/graph/generative-ai-engineering-with-transformers-llms): Adapting and fine-tuning transformer-based LLMs. IBM via Coursera, October 2026.
- [Generative AI Essentials](https://azanello.com/graph/generative-ai-essentials): Generative AI concepts, LLMs and image generation. IBM via Coursera, June 2026.
- [Generative AI Foundational Models for NLP & Language Understanding](https://azanello.com/graph/generative-ai-foundational-models-for-nlp-language-understanding): N-gram, embedding and sequence-to-sequence models for NLP. IBM via Coursera, October 2026.
- [Generative AI Language Modeling with Transformers](https://azanello.com/graph/generative-ai-language-modeling-with-transformers): Transformer-based language models, GPT-style generation and translation. IBM via Coursera, October 2026.
- [Generative AI: Prompt Engineering](https://azanello.com/graph/generative-ai-prompt-engineering): Prompt design techniques for large language models. IBM via Coursera, June 2026.
- [Machine Learning with Python (V2)](https://azanello.com/graph/machine-learning-with-python-v2): Classification, clustering and an introduction to CNNs in Python. IBM via Coursera, October 2026.
- [Multi-Agent Systems with LangGraph](https://azanello.com/graph/multi-agent-systems-with-langgraph): Supervisor and state-machine patterns for multi-agent LangGraph systems. Edureka via Coursera, May 2026.
- [Python Project for AI and Application Development](https://azanello.com/graph/python-project-for-ai-and-application-development): A Python application project built around AI services. IBM via Coursera, June 2026.

## Contact
- Email: me@azanello.com
- GitHub: https://github.com/AleSZanello
- LinkedIn: https://www.linkedin.com/in/alejandrozanello/
- Credly: https://www.credly.com/users/azanello
- X (Twitter): https://x.com/AleSZanello
- Upwork: https://www.upwork.com/freelancers/~01851ff7d016140d3d