Products / Distinct AI

Credit-free, local, AI-powered multi-media software suite in active production and publicly available. Distinct AI is a featured partner of Intel & AMD.

Vector FX (AI Vector Generation)

Convert prompts or images into scalable SVGs for logos, design, and illustration.

Augmentation / n8n

An official n8n Creator doing applied research and hands-on builds focused on n8n workflow augmentation, agentic AI architecture, and multi-model deployment. Key areas include designing secure agentic workflows with privacy-first data routing, benchmarking local vs. cloud inference for token and cost efficiency, and refining context and memory management.

GitHub

Agentic RAG (Official n8n approved)

Uses an LLM-driven agent to plan, route, evaluate, and iteratively refine its retrieval strategy from multiple Supabase databases.

Autonomous AI Customer Support Agent with Multi-KB RAG & Human Escalation (Official n8n approved)

A 50-node production-grade customer support agent that handles inbound support requests end-to-end. It validates and sanitises incoming requests, queries multiple knowledge bases, falls back to web search when the KB returns no result, routes answers by confidence score, and escalates to a human agent when needed (all asynchronously).

Multi-Knowledge Base Agentic RAG Assistant with n8n, Supabase and Ollama (Intermediate)

A production-ready agentic RAG assistant that queries multiple Supabase vector stores simultaneously and routes answers through an AI agent. The agent autonomously selects which knowledge base to search based on the query, returns a grounded answer with source citations, and falls back gracefully when no relevant content is found.

AI Research & Setup

Research Focus:

Figma Design Systems: Build design systems using Design Tokens, Variables, atomic component architecture, and dark/light theme switching; design-to-code structure with WCAG compliance.
Human-in-the-Loop & Agentic Design Patterns: Designed HITL interventions, confidence-based gating, and escalation flows that define when autonomous agents require human oversight and how the handoff surfaces to users.
AI Governance Frameworks: Designed audit infrastructure, PII masking workflows, and compliance pipelines for AI deployments in regulated industries (SOC 2 Type II, PCI-DSS v4.0, SEC/FINRA, OSFI E-23), ensuring AI outputs remain auditable, traceable, and defensible.
Conversational AI Prototypes: Designed and built conversational AI agents with compliance guardrails and trust calibration patterns; published automation templates integrating RAG, multi-agent orchestration, and Supabase vector databases as an Official n8n Creator.

Setup:

IDEs & Code Editors: Cursor, VS Code, Claude Code, Replit (next)
Agent Workflows & Integrations: Claude Code, n8n, MCP (Model Context Protocol) Servers, custom APIs; Google ADK, LangChain/LangGraph (next)
Model Runtimes & Inference APIs: Claude API, Perplexity API, Google Gemini API/ AI Studio (cloud); Ollama, Docker, OpenRouter (local)
AI Models: Claude, Perplexity, Google Gemini (proprietary); Distinct AI’s self-quantized models, and low-parameter/MoE models (Gemma, Cohere, and similar model families)
Prototyping & Ideation: Figma, Lovable, Google AI Studio, Stitch, Nano Banana, Distinct AI
Client UI & Knowledge Management: Open WebUI, Obsidian, Notion, Agentic RAG, RAG, Vector Database, Supabase