SG
Available for new projects

I build AI systems
that don't break.

$ currently shipping AI agents

Seb Gonzales — software engineer working across the stack, from polished interfaces to backend systems and agent orchestration. Turning complex problems into elegant, dependable software.

6+ products shippedAI · Web3 · Full-stack

AI products, automation systems, and shipped builds.

A mix of agentic workflows, user-facing AI experiences, and full-stack products built to be used, not just demoed.

  1. 01
    Diwara AIAI Product

    An AI-powered study companion that lives on your screen. It combines an embedded web assistant, desktop-overlay behavior, and context-aware chat so users can ask questions without breaking focus.

    AI AgentsNext.jsOpenAISpring Physics
  2. 02

    A multi-agent AI system that automates complex workflows. Agents collaborate to research, analyze, and generate reports autonomously. Built with LangChain and CrewAI for intelligent task delegation and execution.

    Multi-AgentLangChainCrewAIPython
  3. 03
    RAG Knowledge BaseKnowledge Systems

    An intelligent document Q&A system using Retrieval-Augmented Generation. Upload any document and chat with it using semantic search, vector embeddings, and LLM-powered responses with source citations.

    RAGVector DBClaude APIEmbeddings
  4. 04

    An AI operations assistant that reads incoming support requests, classifies urgency, proposes replies, and routes work to the correct queue while preserving human oversight.

    AutomationClassificationLLM OpsNode.js
  5. 05
    Research Analyst AgentDecision Support

    A research workflow agent that gathers sources, extracts findings, compares viewpoints, and assembles clean briefings for product, strategy, and growth decisions.

    AgentsWeb SearchSummarizationReporting
  6. 06
    Lead Qualification AgentRevenue Automation

    An AI-driven lead enrichment system that scores prospects, compiles account context, drafts outreach angles, and feeds decisions back into the sales workflow.

    CRMEnrichmentAutomationOpenAI
  7. 07

    A multi-chain and fully on-chain decentralized messaging application that allows users to add, message, block and unblock users all in the blockchain.

    DappChat
    Web3SolidityBlockchainReact
  8. 08
    TradingLeagueProduct Engineering

    A cryptocurrency trading game with 3 game modes (Normal, PvP and Ranked). Practice daytrading in the most enjoyable and interactive way with real-time market data.

    TradingLeague
    ReactNode.jsWebSocketCrypto
  9. 09
    JobbySaaS

    A job-hunting management tool that allows users to add, modify and visualize their job application through charts summarizing their monthly statistics.

    Jobby
    Full-StackChartsMongoDBExpress

The tools I reach for, across the whole stack.

I work where UX, systems design, and AI capability have to fit together — from interfaces down to agent orchestration and the operational habits that keep products reliable.

#Frontend & Product

  • JavaScript
  • TypeScript
  • React
  • Next.js
  • React Native
  • Redux
  • Tailwind CSS
  • Framer Motion
  • Solidity
  • Web3.js
  • Ethers.js
  • Material UI

#Backend & Systems

  • Node.js
  • Express.js
  • Python
  • Ruby on Rails
  • AWS
  • Firebase
  • Google Cloud
  • MongoDB
  • PostgreSQL
  • Redis
  • Docker
  • REST APIs
  • GraphQL

#AI & Automation

  • LLMs
  • Ollama
  • OpenRouter
  • AI Agents
  • RAG Pipelines
  • Vector Databases
  • n8n
  • Multi-Agent Systems
  • AI Workflow Automation
  • AI-Powered DevTools

#Delivery & Tooling

  • Git
  • GitHub
  • CI/CD
  • Jest
  • Postman
  • Figma
  • Vercel
  • TDD
  • Agile

What I'm thinking about right now.

Threads I keep pulling on between projects — where I tend to spend curiosity and side-project time.

01

AI companions

Assistants that stay in context and feel present, without getting in the way of real work.

02

Retrieval pipelines

RAG systems with clean ingestion, semantic search, and answers that cite their sources.

03

Multi-agent workflows

Agents that delegate, supervise, and coordinate tasks instead of just generating text.

04

Quietly sophisticated UX

Product experiences that feel simple even when the system underneath is doing a lot.

Building AI products, automations, or full-stack software?

I work best where thoughtful UX meets solid engineering — especially around AI agents, retrieval, workflow automation, or product velocity.