Repaired and tuned the JQ AI SYSTEMS service and discovery layer: fixed the consulting service route, reordered the service hierarchy around assessment, consulting, lead generation and AI stack setup, synced the homepage preview and service metadata, cut the blog archive from about 765 KB to 79 KB with pagination and leaner schema, reduced broad search pages from about 402 KB to 64 KB, tightened short-query precision
Model-flexible AI systems · Porto · Worldwide
AI Automation Systems.
Your team
should be doing
the interesting work.
Your team is spending hours on work that should take minutes. I build the systems that give those hours back: reporting tools, content pipelines, automation that runs itself. So your people focus on what matters.
10
Anthropic Certifications
17
Systems Built
1,200+
Brands Designed
Now
Studio notes
Built the first working layers of a local voice-cloning studio and B2B voice-outreach lab: FastAPI studio backend, pluggable speech engines, automatic language routing between Qwen3-TTS and Chatterbox, voice-profile transcription support, a Next.js studio UI, batch personalized audio generation from CSV or JSON leads, normalized MP3 outputs, zip downloads, and branded per-lead player pages. The direction stays consen
Shipped a site-wide JQ AI SYSTEMS voice-message widget: visitors can open a small floating panel, hear a short welcome message, record a private audio note from desktop or phone, preview it, and send it with name and email. The flow includes CSRF protection, honeypot handling, upload-size limits, per-IP rate limiting, audio-format checks, private storage outside the public site, email attachments, visitor confirmatio
At a glance
What is JQ AI SYSTEMS?
JQ AI SYSTEMS is an AI automation consultancy in Porto, Portugal, founded by João Queirós, a systems builder with 10 Anthropic certifications and 12 years building brand identities for 1,200+ companies across 40+ countries. He designs and builds model-flexible AI systems: automation pipelines, multi-agent architectures, prompt engineering frameworks, and hands-on setup using the OpenAI stack, including Codex and ChatGPT/GPT models, plus Claude, Gemini, and open-source LLMs when they fit the workflow. Clients range from startups and solo founders to agencies, studios, and established firms across the United States and Europe. Every project starts with a free 30-minute consultation, is quoted fixed price in writing, and is delivered with full source code, prompt libraries, and documentation, so the client owns the system outright. No subscriptions, no lock-in, and a 14-day bug-fix warranty on every build.
Services
How I help
How I help
you move forward.
From one researched decision to a deployed system. No technical knowledge required on your side.
01
AI Problem-to-Solution Assessment
Bring one workflow that wastes time or loses opportunities. I research up to three realistic AI or non-AI options and give you one recommendation plus a 30-day test plan within five business days.
02
AI Consulting & Roadmapping
Not sure where to start? I map your business against what AI can realistically do now. Opportunity identification, priority ranking, and a phased implementation plan. A clear picture of what to build, in what order, and why.
03
AI Lead Generation & Outreach
Fill your pipeline on autopilot. A system that finds prospects from Google Maps, Reddit, and LinkedIn, scores each one for fit with AI, and drafts a personalised first message. Qualified, ready-to-contact leads instead of a blank page.
04
AI Stack Setup & Enablement
Get your whole team productive with AI. I install and configure the right tools (Claude, Codex, Hermes, plus image and video generation), load them with custom skills, then run a hands-on onboarding session so the team uses them well.
05
Pinterest Growth Automation
Reach that compounds while you work. On-brand pins published daily through a zero-cost pipeline (Cloudinary, GitHub Actions, Buffer). Proven on real accounts: 7.1K monthly views in about 4 weeks on one, 66K on another (July 2026).
06
Your workflow could be next.
Bring the repetitive task, messy handoff, or internal AI tool idea that keeps costing time. I will help turn it into a scoped custom build.
Systems
Built and deployed.
Real systems built starting with my own studio, now available for yours.
Published Product · Microsoft Store
Snapline
Capture it. Keep it. Find it.
Screenshots and screen recordings, organized into projects. Local-first: captures save to your own folders, and offline OCR stays on your PC.
Live · Internal
OutreachIQ
Full-cycle outreach pipeline. Google Maps search, Firecrawl + Apollo enrichment, Claude lane scoring, three tone-audited email variants, review queue, send gate, and throttled Gmail sending with outcome tracking. All from one dashboard.
Outcome: 1,600+ real leads processed. 320 emails sent through the review gate under a 30/day cap.
Live · Internal
Tidesstunes Social Pipeline
YouTube music videos converted to 60-second vertical clips and posted across TikTok, Instagram Reels, and Pinterest via Buffer API. 17 videos, 39+ posts queued.
Outcome: 30+ minutes per video to ~2 minutes
Live · Internal
Adobe Stock Uploader
Reads batches of images visually and generates Adobe-ready titles, exactly 46 ordered keywords, asset-level category IDs, normalized JPEG CSV filenames, and validation checks before Adobe Stock Contributor submission.
Outcome: 108 rows validated with 46 unique keywords per asset, valid category IDs, unique filenames, and no banned terms.
Live · Client Delivered
Four Sparrows: Brand to Automation
End-to-end delivery for a Boca Raton social club: brand identity, website direction with a documented WordPress handoff, a playable American mahjong web game, and an autonomous daily Pinterest pipeline.
Outcome: 134 pins posting daily at $0/mo; 7.1K Pinterest monthly views about 4 weeks after launch (July 2026).
Added: 16 Jul 2026
Added: 16 Jul 2026
Live · Internal
Etsy to Pinterest Pipeline
Full-stack FastAPI dashboard with Claude Sonnet niche-aware SEO, smart image processing, perceptual deduplication, and dual posting via Make.com or Pinterest API v5. Replaces 60 minutes of daily cross-posting with a 5-minute review.
Added: 20 April 2026
Live · Public Tool
Codex Control Center
Local-first Codex dashboard for observability, usage insights, workspace health, publishing readiness, and approval-gated tasks. No API key required for local observation.
Outcome: Safer local Codex operations with public demo, video walkthroughs, and GitHub-ready docs.
Studio demos
About
Designer who
Designer who
builds systems.
Twelve years building brand identities. 1,200+ brands across 40+ countries. Then AI changed everything. I didn't switch careers. I added a second layer: building model-flexible automation systems that let creative businesses operate at a scale they couldn't reach before.
The result: systems that think in brand voice, understand design context, and run without babysitting.
Read My Story →2012–2023
Branding Specialist
1,200+ brand identities delivered across 40+ countries. Full portfolio at joaoqueiros.com
2023
JQ AI SYSTEMS Founded
Applied design thinking to AI systems. First custom LLM-powered systems built. First client systems shipped and running in production.
2026
10 Anthropic Certifications
Anthropic credential layer added to a broader stack. Building with the OpenAI stack, including Codex, plus Claude, Gemini, and other LLM tooling as the workflow requires.
Adobe Stock Contributor
25,000 images & 1,500+ videos, AI-generated with Midjourney
Library
From the field.
Experiments, case studies and systems thinking from real builds.
28 Jul 2026 · AI Video
Claude + Higgsfield MCP: Build Vox-Style Motion Explainers From One Prompt
Zubair Trabzada shows Claude Opus 5 orchestrating research, scripts, editorial collage visuals, charts, voice, music, and rendering through Higgsfield MCP. Here is the source-checked workflow, an original production prompt, cost controls, and the human review gates it still needs.
Read Article
28 Jul 2026 · AI Policy
Why Kimi K3 Has OpenAI, Anthropic, and Washington on Edge
Theo argues that Kimi K3 has frontier labs and Washington panicking. Here is the evidence: released open weights, competitive benchmarks, distillation allegations, cyber testing, policy risk, and what builders should do next.
Read Article
28 Jul 2026 · AI Agent Architecture
How to Build a Marketing Agent That Learns From Revenue
Cody Schneider and Greg Isenberg map a real Meta Ads agent: unified business data, research and creative pipelines, bounded API actions, and a feedback loop tied to revenue.
Read Article
--imagined-in midjourney
Where worlds are prompted. JQ VISIONS, my Midjourney gallery.
Open JQ VISIONS →
--prompted-in atlas studio · independent prompt workspace
ai.joaoqueiros.com/atlastudio →
Proof
What clients
What clients
actually say.
Real humans running real businesses on builds that ship. No invented quotes.
Before the build, my Monday morning was a four-hour scramble across every client’s brand, voice, and feed. After João shipped the system, the same briefing lands in thirty minutes and I still recognize my own voice in every draft. I have taken on more clients since, without hiring anyone to cover the repetitive work.
More client testimonials coming as new builds ship.
FAQ
Common questions,
Common questions,
straight answers.
What is JQ AI SYSTEMS?
JQ AI SYSTEMS is an AI automation consultancy founded by João Queirós, a branding specialist with 12+ years of experience and 10 Anthropic certifications. Based in Porto, Portugal, the studio designs and builds model-flexible AI systems: automation pipelines, multi-agent architectures, prompt engineering frameworks, and hands-on setup using the right stack for each workflow, including the OpenAI stack (Codex and ChatGPT/GPT models), Claude, Gemini, and open-source LLMs when appropriate.
What does an AI automation consultant actually do?
An AI automation consultant maps the repetitive, high-volume parts of your business and replaces them with custom AI systems. At JQ AI SYSTEMS: Python pipelines, Codex-assisted builds, multi-agent workflows, scheduled reporting systems, content pipelines, and practical AI setup for teams. Strategy, build, deploy, and handover, all handled end to end.
How much does a custom AI automation system cost?
Custom builds follow two ladders. Quick Builds remove one focused problem: a Sprint (days to one week) starts from $1,035 (€900); a Single-Purpose System (one to two weeks, one full workflow) starts from $2,760 (€2,400). System Builds run a whole operational area: a Mid-Scope System (three to five weeks, scheduled automation or content pipeline with memory) starts from $6,325 (€5,500); a Multi-Agent Platform in the Camille pattern (four to eight weeks, multiple AI agents with memory and orchestration) starts from $13,800 (€12,000); Enterprise Platforms with custom infrastructure and ongoing delivery start from $28,750 (€25,000). Every project begins with a free 30-minute consultation where the highest-impact automation is scoped and quoted precisely before any commitment.
How long does it take to build a custom AI automation?
A focused single-purpose automation (like a reporting tool or scheduled briefing) typically ships in one to two weeks. Full multi-agent systems with memory and client deliverables take four to eight weeks depending on integrations.
What is the difference between Zapier and a custom AI automation system?
Zapier connects apps with simple triggers. A custom LLM-powered system reasons about your data, writes in your brand voice, remembers context across runs, and handles complex multi-step logic that no-code tools cannot. JQ AI SYSTEMS builds the layer above Zapier, not a replacement.
Do I need to be in Porto to work with you?
No. JQ AI SYSTEMS is based in Porto, Portugal, but all work happens remotely. Current clients are in Portugal, France, and the UK. Communication in English, Portuguese, or French.
Do you work with clients in the United States?
Yes. JQ AI SYSTEMS works entirely remotely and takes on companies and startups across the United States. I keep working hours that overlap both US Eastern and Pacific time, run projects asynchronously around your schedule, and hand everything over with full documentation, so being based in Porto, Portugal is never a constraint on US work.
Do I need technical knowledge to work with you?
No. You describe the problem in plain language. JQ AI SYSTEMS handles strategy, design, build, and deployment end to end. You receive a working system plus documentation on how it runs.
Which AI stack do you use for business automation?
The stack depends on the workflow. The OpenAI stack, including Codex and ChatGPT/GPT models, is strong for building, coding, agentic work, and conversational workflows; Claude is excellent for long-context reasoning and brand-sensitive output; Gemini and open-source LLMs can fit other constraints. JQ AI SYSTEMS chooses the model and tooling around the job, not the other way around.
Do I own the code after the project ends?
Yes. You own every line of code, every prompt, every config file. Every build ships with full source and documentation, so you are never locked in for maintenance.
Free Consultation
Tell me your problem.
Tell me your problem.
I map it in 30 minutes.
One call. One honest answer. If it is not a fit, I say so in the first five minutes. No pitch, no pressure, no commitment.
Book Free 30-min Call
I read every message personally
Reply within one business day