One engine.
Purpose-built for the mission.
A workbench is the fusion engine scoped to a mission: the same collection, credibility weighting, and correlation underneath — with the tools, feeds, and overlays the operation actually needs on top. Nothing is a separate product; every workbench inherits the full evidence chain.
Conflict Tracking
The default workbench. Live theatres with events, actors and territory, unit symbology, escalation indicators, and the causal graph — confirmed intelligence on the globe with the full evidence chain, scoped into the battlespaces you define. Replay the picture over time; audit every decision.
Read the workbench manual →Investigations
Work a question to a finding. Threads, hypotheses, and structured analysis panels hold the reasoning; every piece of supporting evidence stays linked to its provenance, so the finding survives the person who wrote it — and the scrutiny that follows.
Read the workbench manual →Analytics
Interrogate any pipeline and publish what you find. Charts, analytical modules, and a report engine that carries the evidence chain into the final document — so the report you hand up is backed by everything Krataxis collated and corroborated to produce it.
Read the workbench manual →Health & Disease
Outbreak tracking and case management on the same evidence discipline — down to scenarios most tools never consider, like tracking an outbreak aboard a moving vessel. Cases, contacts, and spread as a corroborated picture rather than a spreadsheet.
Read the workbench manual →Disaster
The same engine, pivoted to humanitarian assistance and disaster response: the event timeline, impact extents, ingress and egress planning, and ground-truth reporting from crews in the field — fused down to the level of a single team.
Full capability set ↓From first alert to the last team out.
Disasters break the assumptions every other tool relies on. Roads flood, airstrips close, comms degrade — and the map from yesterday is a rumour. The Disaster workbench treats a crisis the way Krataxis treats everything: as an evidence problem. What is confirmed, what is corroborated, and what can the next crew in actually rely on.
No feed has a clean "disaster" category, so Krataxis builds one: a custom aggregate scope across GDELT's adjacent themes — humanitarian response, aid provision, displacement — fused with GDACS event alerts. The coverage is assembled deliberately rather than inherited from another provider's taxonomy.
The causative graph builds the disaster's timeline as it unfolds and resolves its physical extents — humanitarian impact hulls, flood and fire polygons — so planners see where the event actually is, not where the first headline said it was.
After a cyclone you can't assume the road network or the airport. The workbench plans maritime ingress and egress alongside ground routes, and replans as the picture changes.
The operating picture fuses down to the level of a single crew: their routes, their reporting, their piece of the AO. Headquarters and the ground see the same evidence, at the fidelity each needs.
The first crew in maps the way for everyone behind them.
Ground truth in a disaster is perishable. When the first team reaches a washed-out bridge, that observation shouldn't die in a radio log — it updates the route network for every following crew, instantly. Four states, one shared picture.
Report once. Reroute everyone.
Each flag carries its provenance like everything else in Krataxis — who reported it, when, and what they saw. Following teams get replanned ingress and egress automatically; command sees the road network degrade and recover in real time.
BLOCKED — impassable pending clearance
BLOCKED — WORKABLE — passable with capability caveat
WORKABLE / OPEN — confirmed passable
From "person overboard" to a waypoint list.
When someone is in the water, the search area is a physics problem with a clock on it. The SAR module turns a rough last-known position into an executable search plan.
Mark the last known position
Rough coordinates and time overboard — that's all the operator needs to supply.
Compute the drift
Local wind, wave, and current data drive a drift model that projects the likely track and expanding probability area.
Optimise the pattern
The system fits the most effective search pattern to the highest-probability area for the assets available.
Hand crews their waypoints
Field teams receive an ordered waypoint list they can fly or steer directly, with no heat map left to interpret.
The same discipline applies here as everywhere else in the platform: the plan shows its working. Every waypoint traces back to the drift model's inputs, so the search coordinator can adjust the pattern the moment conditions change.
Put your operation on the workbench.
The Disaster workbench is in final deployment. Relief organisations, rescue services, and HADR planners can register interest now — early partners shape the toolset.