Krystal Unity builds focused AI products and the operating layer behind them: threat intelligence, blockchain evidence, voice workflows, field diagnostics, analytics, and runtime governance. The portfolio mixes live products, beta work, and paused experiments; the shared direction is MCP-connected, auditable agent infrastructure.
A live and emerging portfolio built around governed, MCP-connected AI workflows.
Deception-first threat intelligence for capturing attacker behavior in controlled environments, preserving evidence, and turning incidents into usable security signals. MCP access is being baked into the roadmap so agents can query detections, evidence, and Sentinel outputs through a governed interface.
blackdome.ai →Blockchain evidence explorer for legal, personal, and investigative workflows. The core product is live-stage; the next major step is MCP access for agentic blockchain analysis, trace pack creation, and repeatable evidence review.
krystaltrace.com →Voice agent platform for high-volume service workflows. Answers calls, books appointments, handles changes and cancellations, and escalates where staff need to take over. Current verticals: Reception, Clinic, Hospital, and Hotel.
Explore Krystal Voice →Voice AI + RAG diagnostics for field service teams. Technicians call in, describe the issue, and FixLine searches approved manuals and SOPs to help identify likely fixes by phone, text, or photo. Built for repeatable service workflows where documentation already exists.
usefixline.com →Runtime control plane for autonomous AI agents. Intercepts every tool call, applies policy before execution, and records auditable decision trails for human review. Designed for regulated teams that need visibility, vetoes, and runtime control around agent behavior.
haltstate.ai →Session analytics for finding user friction through replays, funnels, heatmaps, and AI-assisted review. Current work is shifting toward broad MCP connectivity so product teams and agents can query behavioral evidence directly.
krystalview.com →Browser-based cybersecurity training spun out of the BlackDome honeypot work. Red-team and blue-team scenarios are mostly developed and currently need hardening, polish, and controlled release work before broad public use.
arena.krystalunity.com →AI-powered grant discovery and application management. Finds relevant grants, tracks deadlines, manages compliance, and automates follow-ups — so you never miss funding you qualify for.
Explore Grant Hunter →MYOB-focused financial insight product built around cash-flow trends and small business forecasting. Development is paused while the market shifts toward native accounting-platform agents and MCP-based finance access.
cashflowtrends.ai →The shared KU stack gives each product a common operating layer for customers, evidence, support workflows, governed access, and product operations.
How it fits together →The portfolio is backed by pending architecture around safe, auditable, governed agentic AI.
Autonomous systems should stay inside human authority, with clear approval and review patterns before important actions happen.
AI workflows need durable context and accountability across tools, sessions, teams, and model providers.
The same governance direction runs through the product stack: evidence, visibility, operator control, and practical deployment.
For product access, pilots, MCP integrations, partnerships, or patent-pending governance discussions.