Applied AI & Algorithmic Research

Intelligent systems for
an algorithmic world

Shikigami Technologies builds AI systems that give individuals and small organizations the operational capability of entities ten times their size. Founded in Puerto Rico. SDVOSB certified. Built for government and commercial clients who need intelligence that works — deployed locally, never in the cloud.

Explore the platform Read our research
Document Intelligence Contract Analysis Vendor Risk Assessment Algorithmic Trading Federal Contracting Local Deployment SDVOSB Certified Puerto Rico AI Automation
Document Intelligence Contract Analysis Vendor Risk Assessment Algorithmic Trading Federal Contracting Local Deployment SDVOSB Certified Puerto Rico AI Automation
01 — What we do

One firm, the full stack of applied intelligence

AI Systems

Design and deployment of production machine-learning systems — from model research to inference infrastructure that scales.

Financial Algorithms

Research, backtesting, and live deployment of trading and risk algorithms across asset classes.

Business Automation

Intelligent workflows that remove manual work — document processing, decisioning, and operations at scale.

Data Analytics

Pipelines and dashboards that turn raw, high-frequency data into decisions you can act on.

Technology Consulting

Embedded expertise to architect, ship, and de-risk complex technology and AI initiatives.

Digital Products

Full-lifecycle software and digital export services — designed, built, and commercialized in-house.

02 — Research

Research in progress

We publish findings when they are production-tested, not before.

03 — Platform

Products built on one intelligence core

A shared foundation of models, data pipelines, and execution infrastructure — surfaced as focused products for research, trading, and automation teams.

ACTIVE — Available now
Cypher
Document Intelligence Platform

AI-powered invoice and document processing deployed locally on client hardware. Cypher reads invoice PDFs, extracts structured data via Azure Document Intelligence, matches purchase orders through cross-document verification, flags fraud and anomalies across six detection rules, and generates Excel master workbooks and PDF audit summaries. Zero cloud dependency — all data stays on the client's machine.

Azure OCR + GPT-4o extraction pipeline
Automated PO cross-document matching
Fraud and anomaly detection (6 rules)
Local SQLite + Excel output, no cloud required
Deployed: active client, 30-location operation
COMING SOON
Ring of Power
Contract Technical Intelligence

The first AI system built for the execution side of federal contracting — not the proposal side. Drop in any government SOW, PWS, or RFP and receive a complete technical analysis: hidden requirement detection, infrastructure and API mapping, realistic execution timeline with UAT buffers, cash flow risk scoring, and a Go/No-Go recommendation matched to your specific capability profile.

Hidden requirement detection (implied compliance flags)
Infrastructure + AI stack analysis
Execution timeline with federal payment reality built in
Go/No-Go scoring vs contractor capability profile
Government and commercial contract tracks
COMING SOON
Anor
Contractor Viability Intelligence

Pre-award vendor risk assessment for federal contracting officers. Before signing the award, Anor cross-references the contract's technical requirements against the vendor's demonstrated capabilities, financial survivability indicators, certification compliance, and past performance patterns from public contract data. Surfaces the vendors that will fail before they fail.

Technical capability gap analysis vs contract scope
Financial viability scoring (cash flow survival modeling)
Certification and compliance verification
Past performance pattern from public contract history
Award risk score with conditional award framework
INTERNAL — Proprietary system
Sentient
Algorithmic Market Intelligence

Shikigami's proprietary algorithmic trading and signal generation system. Operates across prediction markets with real-time regime detection using Markov state analysis, oracle lag scanning, order book imbalance detection, and funding rate contrarian signals. Six supervised engines running continuously with automated restart, watchdog monitoring, and Telegram-based human approval for all live executions.

Markov regime detection (5-state FNG routing)
Multi-signal confluence: oracle lag, CVD, CLOB imbalance
Six supervised engines with auto-restart
Quarter Kelly position sizing with regime-adjusted floors
Shadow mode gate: 50 clean trades before live capital
04 — Academy

Knowledge base coming soon

Documentation and technical guides for deployed Shikigami systems.

Let's build what's next in intelligence

Partner with us on a deployment, ask about Ring of Power or Anor, or tell us what you're working on.

hello@shikigamitechnologies.com