Due diligence · AI governance · Cost control

The scorecard PE firms wish they had before close.

A sovereign, fully local solution to help private equity firms measuring their AI & data exposure. Scored from real signals inside the business in under 24 hours, not declarative questionnaires. It becomes a companion that helps portfolio companies improve their spent, usage, after the deal.

AI & Data Due Diligence · Portfolio Co.
report ready
B
74 / 100
✓ 3 passed ⚠ 4 watch ✗ 2 risk ↑ 6 boosted
Regulatory Risk
computed
AI Guardrail Coverage
booster
Vendor Concentration
booster
Shadow AI Exposure
booster
AI Data Leakage Prevention
sub-engine
+ 6 more KPIs · expand for full breakdown & evidence

Two moments, one platform

Built for the deal and the decade after.

Most AI audits are a snapshot in time. We give you a live instrument panel that stays relevant long after the deal closes.

📋

Pre-close due diligence

Plug in during the DD process. In under 24 hours you get a scored, cited AI & data readiness report covering areas no traditional advisor currently touches. Exportable as a self-contained report ready for the IC memo.

🧭

Post-close companion

The platform stays on as a live companion for portfolio companies. Sub-engines called Boosters actively help teams improve each KPI so scores move with the business, not just at audit time.


Coverage

11 KPIs that don't appear in traditional audits.

Every score is computed from real signals inside the business not self-reported surveys. Each KPI carries a status and an active Booster that TorvexLabs has developed to help you close the gap.

Regulatory Risk
Your regulatory risk and data residency exposure scored from your systems and data signals, not from policy documents or questionnaires.
Vendor Concentration
Dependency on a single LLM, infrastructure, or data pipeline provider. Includes whether routing diversification is in place.
↑ booster available
Governance Maturity
AI usage policies, approval workflows, and accountability chains. Includes whether human-in-the-loop governance is formalised.
AI Dependency Risk
Model lock-in exposure across projects. Includes whether routing abstraction and fallback models are operational.
↑ booster available
Lifecycle Hygiene
Stale assets, archive discipline, and AI-generated artifact sprawl: the hidden drag on engineering velocity.
↑ booster available
AI Guardrail Coverage
Prompt-injection exposure, sensitive data leakage detection, and enforcement of output policies at runtime.
↑ booster available
Shadow AI Exposure
Untracked AI tool usage across the org: personal API keys, ungoverned SaaS, and unregistered model access.
↑ booster available
AI Data Leakage Prevention
Where data moves, what is classified, and whether data loss prevention is in place and verified end-to-end.
↑ booster available
AI Incident Response
Runbooks, SLAs, drills, and post-mortem discipline for AI-related failures. Scored before they become investor calls.
↑ booster available
AI Observability Maturity
Logging coverage, monitoring pipelines, and audit trails for AI workloads: what you can see and what you can prove.
Environmental AI Governance
Carbon and energy footprint of AI workloads: tracking, reporting, and reduction targets per project and per model family.

Live companion

AI usage tracked, optimised, and explained on an ongoing basis.

Not a one-time audit. The companion dashboard tracks token consumption, AI costs, carbon footprint, and project-level efficiency every day.

When a KPI dips, a Booster surfaces giving users actionable guidance inside the tool they're already using, rather than a red flag in a report no one re-reads.

AI Usage Driving Seat
7d 30d YTD All
AI Spend
$8,240
Total tokens
2.41M
CO₂ (kg)
12.4
Energy (kWh)
44.2
Carbon footprint ≈ 73 km driving · 3,210 phone charges · 2.4 FR household-days
Project Sessions Cost CO₂ Intensity
Project 1 312 $2,841 4.2 kg heavy
Project 2 187 $1,920 2.8 kg medium
Project 3 94 $780 1.1 kg light

Carbon Footprint · Sustainability

Every AI request has a carbon cost. We make it visible.

AI Command & Control tracks CO₂ per model, per project, and per team in real time. The AI Router doesn't just optimise for performance it routes to EU-sovereign, lower-carbon models when tasks allow, reducing your portfolio's footprint without sacrificing output quality.

Compare models by carbon intensity, see where heavy compute concentrates, and report green AI progress to LPs and boards with confidence.

CO₂ Footprint · Portfolio Co.
● tracking
847g
CO₂ this week
↓ 18%
23.4 kg
saved via AI routing
≈ 73 km
driving equivalent
CO₂ / 1,000 tokens · by model
Mistral 7B
0.8g EU ✓
Llama 3 EU
0.9g EU ✓
Claude Haiku
1.1g
GPT-4o mini
1.4g
Claude Sonnet
1.8g
GPT-4 Turbo
2.3g heavy

AI-specific analyses · 5 modules

Go deeper on what matters. Five targeted analyses.

Beyond the KPI score, each deep-dive module runs a dedicated analysis on a specific AI risk area pulling signals from your systems, testing coverage gaps, and ranking findings by severity.

Each module surfaces findings inside the platform and auto-assigns a Booster when there's a clear remediation path. All 5 run in parallel, orchestrated by Claude.

C
Claude · Portfolio Analysis
● active
Run full deep-dive suite on Portfolio Co.
Prioritise by risk score.
C
Running all 5 modules ranked by risk materiality.
01 Shadow AI Detection ▶ running
02 Vendor Sovereignty Audit ▶ running
03 Prompt Injection Check ● queued
04 AI Governance Review ● queued
05 Data Leakage Exposure ● queued

AI Map · Sovereignty layer

Know exactly where your data goes and who controls it.

The AI Map auto-discovers every connected service, model provider, and data flow. It plots them against two axes: architecture layer (AI Apps → Cloud → Data) and sovereignty zone (EU sovereign vs non-EU dependency).

The AI Router the heart of the system shows which model each project actually routes through, and whether that path crosses a sovereignty boundary. Mistral stays in-zone; OpenAI doesn't.

EU · SOVEREIGN · PROJECTS · NON-EU DEPENDENCY AI APPS CLOUD SYSTEMS DATA routed adhoc Mistral EU ✓ AI Router OpenAI US ⚠ Scaleway EU ✓ Project 1 Project 2 AWS US ⚠ Postgres EU ✓ Project 3 EU sovereign Non-EU dependency Project AI Router adhoc

Pre-deal · Due diligence report

A scored report ready for the IC memo. In under 24 hours.

Connect during the DD process. Within 24 hours you receive a structured AI & data readiness report scored across 11 KPIs, cited with evidence, and ranked by materiality for the deal.

Exportable as a self-contained PDF. Covers areas no standard tech advisor currently touches, with remediation priorities included ready to hand to the portfolio company on day one.

AI & Data Due Diligence · Portfolio Co.
report ready
B
74 / 100
✓ 3 passed ⚠ 4 watch ✗ 2 risk ↑ 6 boosted
Regulatory Risk
computed
Governance Maturity
computed
AI Guardrail Coverage
booster
Vendor Concentration
booster
AI Dependency Risk
booster
Shadow AI Exposure
booster
AI Data Leakage Prev.
sub-engine
+ 4 more KPIs · export full report as PDF

For investors

The gap traditional tech DD leaves open.

AI & data readiness is a core value driver in every portfolio company today. It's invisible in most investment memos. We change that.

11
KPIs not covered in standard tech due diligence
<24h
Time from connection to first scored report
100%
Computed from real signals no self-reported surveys

See your score and pilot AI efficiency, minimizing risks.

Connect your stack. Get a scored, cited AI & data report. Keep it live as a companion and a competitive edge after the deal.

Request early access