01 / FRAME
Find the actual decision.
Separate symptoms from constraints. Map stakeholders, incentives, evidence, uncertainty, and where human judgment still matters.
Business development / service architecture / AI systems
From messy problems
to working systems.
I’m Janko Štrkalj. I turn ambiguous, high-stakes problems into systems people can actually act on: across AI, risk, security, and enterprise workflows.
janko / current_focus.yaml
# working at the seam between
problem: "too much signal, not enough decision"
current_builds: [Riskform, GRC Assistant, Evidence Manager]
domains: [AI, risk, security, GRC]
method: frame → model → build → test
bias: explainable systems over black boxes
status: shipping
How I work
I like the part before the obvious solution exists.
The useful work is usually not adding another dashboard. It is finding the decision hiding underneath the workflow, making the logic explicit, and designing a system that survives contact with reality.
02 / MODEL
Make the logic inspectable.
Turn tacit reasoning into policies, states, workflows, thresholds, and escalation paths that can be challenged and improved.
03 / SHIP
Build the smallest credible system.
Prototype the interface and operating model together. Test whether the decision gets faster, clearer, and easier to defend.
Current builds
Riskform / Risk Reduction Platform
Stop prioritizing vulnerabilities. Prioritize the work.
Riskform turns fragmented vulnerability and threat intelligence into explainable remediation campaigns: ordered by the risk they can remove and the effort required to remove it.
public data first
policy as code
explainable decisions
campaigns, not CVEs
PRIVATE BUILD / NOT PUBLIC YET
INPUT
NVD · CVSS · EPSS · KEV · threat intel
signals
↓
POLICY
taxonomy · scoring · grouping · planning
explain
↓
OUTPUT
remediation campaigns
now / next / later
GRC Assistant / Private Build
Put AI inside the workflow, not on top of it.
GRC Assistant explores how cybersecurity maturity, assessments, controls, evidence, risk, and audit readiness can operate as one connected system, with humans retaining decision authority.
human in the loop
structured evidence
compliance as code
AI-native workflows
PRIVATE BUILD / NOT PUBLIC YET
MODEL
Maturity → Assessment → Controls
structure
↓
EVIDENCE
documents · validation · linked risks
context
↓
DECISION
risk · mitigation · audit readiness
human-led
Evidence Manager / Private Build
Turn scattered files into connected, searchable context.
Evidence Manager is a private, AI-assisted workspace for organizing files, notes, references, and saved material into logical objects that can be searched, connected, and revisited by meaning instead of folder location.
local-first
semantic search
AI-assisted linking
personal knowledge
PRIVATE BUILD / NOT PUBLIC YET
INPUT
files · notes · links · screenshots · references
collect
↓
AI LAYER
extract · relate · cluster · describe
connect
↓
OBJECT
searchable evidence with context
retrieve
A recurring theme
AI is most useful when it becomes infrastructure for judgment, not decoration around a workflow.
Across Riskform, GRC Assistant, and Evidence Manager, the interesting question is not “where do we put the chatbot?” It is how to structure evidence, context, dependencies, and human checkpoints so better decisions happen inside the workflow.
The common principle is simple: keep evidence attached to context, make relationships explicit, and reduce a large information space into something people can actually navigate and act on.
Proof of work
Less personal brand. More working surface.
What I’m building, how I think, and where to reach me.
PRIVATE BUILDRiskform
Risk reduction and remediation planning.
PRIVATE BUILDGRC Assistant
Human-led AI workflows for GRC.
PRIVATE BUILDEvidence Manager
Connected, searchable context for files and references.
WORK / CONTACTLinkedIn
↗
Open loop
Working on a decision-heavy system?
Security, GRC, AI-assisted operations, enterprise workflows: if the problem is messy, cross-functional, and hard to make operational, that is usually the interesting part.