Current pilot focus
Automotive HARA chain and TARA
Pilot workflows focus on project context, Item Definition, HARA, Safety Goals, TARA, review queues, traceability, source grounding, and evidence readiness.
AI Safety Engineering Platform
Accelerating Regulated SystemsFrom Concept to Compliance
Platform walkthrough
See how a safety workflow moves through the platform, from project setup to traceable outputs.
Aegis SafeForge helps regulated engineering teams move from project context to reviewed HARA, TARA, requirements, goals, evidence, and traceability while keeping engineers in control.
AI assists with structured drafts.
SafeForge governs review, traceability, approvals, and change impact.
Engineers make the final decisions.
ISO 26262, 21434, 62278
Multi-Industry
Platform Core
Active development
Advanced safety and industry artifacts
FMEA, FMEDA, FTA, ETA, HAZOP, LOPA, lifecycle dashboards, expanded reporting, and industry-specific artifacts are being developed and refined with design-partner feedback.
Expansion tracks
Railway, medical-device, and process safety
Additional regulated engineering profiles are being extended through configured pilots and domain-specific workflow configuration.
Concept to compliance
Compliance readiness should not start at the end of a project
SafeForge is being built to help teams identify missing evidence, weak traceability, stale decisions, and open review tasks earlier in the engineering lifecycle. Instead of preparing disconnected compliance documents at the end, teams can turn scope assessments, risk decisions, requirements, controls, approvals, and evidence into governed platform artifacts as work progresses.
Aegis SafeForge does not replace legal advice, certification bodies, or conformity assessment. The platform is being developed with human oversight, traceability, cybersecurity, audit logs, source grounding, and controlled AI assistance as core design principles.
Discuss a compliance workflowRegulatory scoping first
Start by identifying which regulations, standards, and customer obligations may apply before teams commit to a safety, cybersecurity, or evidence workflow.
Gap and readiness visibility
Track missing evidence, weak links, stale decisions, unreviewed artifacts, and open ownership before those gaps appear during expensive reviews or audits.
Documentation as governed artifacts
Maintain scope assessments, obligation mappings, risk controls, evidence bundles, review records, and export packages as part of the working process.
Compliance-by-design platform thinking
SafeForge itself is being formalized around human oversight, traceability, audit logs, access control, cybersecurity, source grounding, and controlled AI assistance.
What You Get
Everything in one connected workspace
SafeForge brings project setup, AI-assisted analysis, review control, and follow-on outputs into one product for safety teams.
Industry-aware workspaces for safety and security workflows
AI-assisted drafts with clear review status
Document parsing and source-grounded retrieval
Item Definitions, HARA, TARA, goals, requirements, claims, controls, and evidence workflows
Manager review queues, ownership, and role-based access
Traceability and change-impact visibility as work evolves
Platform Capabilities
What drives the workflow
SafeForge combines controlled drafting, review visibility, evidence grounding, and linked outputs inside one industry-aware workspace.
01
Project Context and Source Material
Start from project definitions, uploaded documents, requirements briefs, approved artifacts, and shared engineering knowledge so the team works from a consistent context.
02
Structured Analysis Workflows
Build and review the pilot-ready automotive workflows around Item Definition, HARA, Safety Goals, TARA, FSR, TSR, and Verification Evidence. FMEA, FMEDA, FTA, ETA, HAZOP, LOPA, hazard-log, and industry-specific risk-analysis profiles are active development tracks.
03
Follow-On Outputs
Carry approved decisions into goals, requirements, controls, claims, validation records, incident-response plans, evidence bundles, and related outputs.
04
Management Oversight
Give managers visibility into review queues, assignments, approvals, stale artifacts, access control, and team responsibilities without leaving the platform.
Why SafeForge
Why Choose Aegis SafeForge?
Regulated engineering teams need more than faster document generation. They need safety and cybersecurity work that moves faster, stays reviewable, and remains defensible as projects change.
Accelerate safety work without losing engineering control
AI helps teams move from project context to first-pass HARA, TARA, requirements, goals, controls, and evidence suggestions faster, while engineers remain responsible for review and approval.
Build traceability into every engineering decision
SafeForge connects item definitions, hazards, threats, goals, requirements, controls, claims, evidence, and approvals during the workflow instead of forcing teams to reconstruct links later.
Keep every approval audit-ready
Review states, comments, ownership, version history, and approval decisions remain part of the engineering record, making it clearer who accepted what and why.
Catch compliance gaps before they become audit issues
Source grounding, traceability, review history, and evidence readiness help teams identify missing context or weak links before reviews, audits, or customer delivery.
Surface change impact before it becomes project risk
When upstream context changes, affected downstream artifacts can be flagged for review so teams can address stale decisions before they become project risks.
Accelerate reviews with structured, auditable workflows
Structured artifacts, reviewer queues, field-level ownership, and clear state transitions reduce the hidden back-and-forth common in spreadsheets, documents, and email threads.
Turn approved work into reusable engineering knowledge
Approved artifacts, source references, decisions, and review history become reusable project context instead of one-off document output.
Fit the right data boundary
SafeForge can support different deployment discussions, from managed SaaS to regional hosting, private cloud, VPC, or self-hosted setups depending on customer security needs.
Built from the problem
Built from real safety-engineering pain points
SafeForge is early, focused, and technically grounded. The product direction comes from functional safety research, founder-led engineering experience, and conversations with safety and cybersecurity professionals.
Originated from University of Kassel master's thesis work and practical safety-engineering frustration.
Focused first on the automotive HARA chain, including FSR, TSR, Verification Evidence, TARA, review workflows, traceability, and evidence readiness.
Designed around human approval, source grounding, deterministic checks, and review control instead of generic AI text generation.
Currently shaped through pilot and design-partner conversations rather than broad unsupported claims.
Supported Industries
CURRENT COVERAGE, WITH ROOM TO GROW
SafeForge is most mature today around automotive safety and cybersecurity pilot workflows. Additional domain profiles are being extended through configured pilots and industry-specific workflow models.
Automotive
ISO 26262, ISO/SAE 21434Functional safety and cybersecurity workflows for road-vehicle programs, starting with Item Definition, HARA, Safety Goals, TARA, review control, traceability, and evidence readiness.
Item DefinitionHARASafety GoalsTARA
Railway
IEC 62278, IEC 62425, CLC/TS 50701Railway RAMS and cybersecurity profiles are being extended through pilots, with domain terminology and outputs designed to fit rail programs rather than forcing a generic template.
System DefinitionSafety Risk AnalysisHazard LogCybersecurity Risk Assessment
Medical Devices
ISO 14971, IEC 62304Medical-device risk-management profiles are being extended through pilots, with guided device context, structured review, and linked risk-control outputs.
Device DefinitionRisk AnalysisRisk-Control OutputsReview Workflow
Process Safety
IEC 61511Process-safety profiles are being extended through pilots for failure analysis, risk reduction, and SIL-oriented review in one governed workspace.
FMEASIL ReviewRisk ReductionApproval Workflow
Platform model
AI proposes the draft. SafeForge preserves the decision path.
Source-grounded suggestions
Deterministic safety logic
Human approval states
Background audit trail
Deployment options
THE RIGHT DATA BOUNDARY FOR YOUR TEAM
SafeForge can be delivered as managed SaaS or in a more controlled environment. We align hosting, region, and operating model with your security and compliance requirements.
Regional availability, data residency, service scope, and technical architecture are confirmed during the deployment fit review.
Discuss deploymentManaged
Cloud SaaS
Aegis SafeForge-operated service for faster onboarding, managed updates, and a maintained production environment.
- Managed application and updates
- Team and organisation access
- Pilot and production options
Regional
Region-specific SaaS
Organisations can request a preferred hosting region. EU companies can request EU servers for their SaaS environment.
- EU servers available on request
- Region agreed during onboarding
- Data-residency needs reviewed
Controlled
Private or self-hosted
For sensitive engineering environments, we support private cloud, private VPC, and self-hosted deployment models.
- Private cloud or VPC
- Self-hosted option
- Air-gap requirements reviewed
Archive
Latest insights
Discuss a pilot
Turn safety and cybersecurity work into traceable decisions
See how SafeForge brings item context, documents, AI-assisted HARA/TARA and requirements drafts, review decisions, and traceability into one governed engineering workflow.
Newsletter
Safety workflow updates
Get product notes, event invites, and practical safety engineering articles when we publish them.