Built by David Daddi

Tools & Systems

From a proprietary multi-agent AI operating system to open-source PPC tooling and a SaaS listing optimizer — built at the intersection of Amazon FBA, AI engineering, and capital operations.

ABOS / FirmAIn

Proprietary · Internal

Multi-Agent AI Operating System for Amazon FBA

ABOS — now operating under the internal name FirmAIn — is a proprietary, event-driven multi-agent AI system built to manage Amazon FBA brands with near-zero manual overhead. It goes well beyond a listing audit tool: it's a full operating system for brand intelligence, operational automation, and capital efficiency.

Rather than a single generalist AI, the architecture organizes specialized agents into three independent squadrons, each with a narrow mandate and the ability to run concurrently:

Operational AI

Handles high-frequency, day-to-day brand operations: inventory forecasting, pricing elasticity testing, and PPC capital allocation — each managed by a dedicated agent, not a single general-purpose model.

Strategic AI

Macro-intelligence layer for portfolio governance and capital yield arbitrage across a multi-SKU, multi-brand ecosystem. Monitors marginal return on capital in real time and reallocates accordingly.

M&A AI

Agents dedicated to acquisition pipeline management, financial due diligence, and deal flow analysis — collapsing the traditional due diligence timeline from weeks to hours.

A built-in Human-in-the-Loop permission engine governs every action. Each potential decision is scored against three profiles: allow (auto-executed), ask (held for human approval via IM), or deny (hard-blocked). No high-impact action executes without validation — which is the exact opposite of the "set and forget" AI most tools offer.

ABOS is used in production to operate David's own supplement brand on Amazon — the system wasn't built as a theoretical architecture, but as the operational backbone of a real brand with real SKUs, real inventory, and real capital at stake.

Connection to the listing service The AI-Native Readiness Score used in every Agentic FBA rebuild audit is one component of this system. ABOS is also the technical foundation informing every feature of Keoxs AIO — the public SaaS currently in development. The architecture was built first; the product is its public expression.

Keoxs Co-pilot

Open Source

PPC Management Tool for Amazon Sellers

View on GitHub

Keoxs Co-pilot is an open-source PPC management tool built for Amazon sellers. It's designed to simplify campaign structure, reduce wasted ad spend, and make Amazon PPC more manageable for small and medium FBA brands that don't have a full-time ads team.

The tool is available on GitHub — free to use, modify, and contribute to.

Open Source Amazon PPC Seller Tools

Keoxs AIO

SaaS · In Development

The Amazon Listing Optimizer Built on Amazon's Own AI Research

Explore Keoxs AIO →

Most listing tools are still fighting the keyword ranking war. Alexa for Shopping — Amazon's AI discovery and recommendation layer — doesn't index keyword density. It builds a semantic model of your product and decides whether to surface it when a shopper asks a question in natural language. Keoxs AIO is built specifically to win on that layer.

The platform is founded on Amazon's own published AI research: COSMO, Amazon's knowledge graph that maps the semantic relationships between products and latent buyer intent; and Amazon's AI shopping framework, which defines the buyer intent dimensions Alexa for Shopping evaluates when deciding what to recommend. No other Amazon SaaS is built on this research — most sellers don't know it exists.

Diagnose
Know your AI Visibility Score

Get a precise, composite score on every ASIN in your catalog — semantic coverage, COSMO knowledge graph mapping, buyer intent gaps, and a simulation of how Alexa for Shopping would actually respond to real shopper questions about your product.

Outrank
Reverse-engineer your competition

Map the full semantic structure of any competitor ASIN. See exactly which buyer intent angles they're covering — and which they're leaving completely undefended. Build your content strategy on their blindspots before they find them.

Optimize
Fix every signal the AI reads

22 targeted tools covering title noun-phrase structure, RAG-ready bullet extraction, backend attribute fill rate, FAQ seeding, image semantic coherence, A+ content analysis, and review-to-listing alignment. Each tool targets a specific layer of AI readability.

Track
Know the moment something shifts

Monitor AI visibility scores over time across your entire catalog. Get alerted before listing drift costs you recommendations — whether it's Amazon auto-editing your content, a competitor improving, or a seasonal relevance shift.

Behind every audit runs an 8-agent AI pipeline that combines three specialized models: Claude handles listing rewrite and compliance validation; Gemini handles visual intelligence, analyzing all seven product images for semantic coherence with the text; and Llama handles semantic scoring, review mining, and FAQ pair generation. No single model owns the whole task — each handles the layer it was built for.

22 Tools total
14 Live now
8 AI agents
3 LLMs
Claude · Gemini · Llama
$49 Starting / month
BSA-Compliant Built exclusively on Amazon's official SP-API — no scraping, no grey-zone workarounds, no fake simulation of Amazon's internal systems. Every feature is fully compliant with the Business Solutions Agreement, including how competitor data is accessed and how automated actions are governed.
Free Forensic Audit — no account needed

Enter any ASIN and get your AI Visibility Score instantly, with the top semantic gaps identified. See exactly what Alexa for Shopping can and cannot extract from your listing today.

Try Free →

Want a hands-on rebuild?

The tools are self-serve. If you want the full AI-Native Listing Rebuild done for your hero product — scored, rebuilt, and guaranteed — apply for your audit.