DWT-Ready October 2026
Simpler Recycling Compliant
AI Governance Transparent
GDPR Article 25
Hospitality Intelligence Platform · Beta Pilot 2026

Hospitality doesn't have
a waste problem.
It has an interpretation problem.

Varensi™ turns kitchen waste signals into recovered margin, compliance confidence, and audit-ready carbon reporting — built for the DEFRA DWT mandate and beyond.

3-Layer Variance Framework™
DEFRA DWT October 2026
GHG Protocol Scope 3
UK AI Governance
GDPR · EU-West-2 London
About Varensi

The intelligence layer above basic waste tracking

Most waste management tools count waste. Varensi interprets it — classifying every record into one of three layers so operators know not just how much waste they're generating, but why, and what to do about it.

"Waste is not a cost to be logged.
It is a signal to be interpreted."
The Varensi Intelligence Philosophy · 3-Layer Variance Framework™
🧠
Automatic AI Classification
Every waste record is automatically classified into the 3-Layer Variance Framework using the Varensi Intelligence Engine. Confidence scores and plain-English reasons are stored for every decision.
🌿
Carbon Intelligence
CO₂e calculated automatically on every record using DEFRA emission factors. GHG Protocol Scope 3 Category 5 aligned. AI emissions tracked and disclosed separately under Scope 3 Category 1.
📋
DWT Compliance Preparation
Interactive pre-submission checklist, responsibility matrix, EWC code assignment, and DWT-ready CSV export. Built specifically for the DEFRA Digital Waste Tracking mandate of October 2026.
💷
Financial Recovery Identification
Systemic waste patterns are costed in GBP using real ingredient prices. The recoverable margin — what could be saved by fixing the pattern — is surfaced on every dashboard view.
🔒
AI Governance Transparency
Every AI classification is explainable, auditable, and subject to human override with a documented reason. Built in compliance with UK DSIT AI governance requirements and EU AI Act transparency obligations.
🎙️
Photo and Voice Entry
Kitchen staff can record waste by photographing a bin or speaking a description. The intelligence engine auto-fills the form from image or voice transcript — reducing friction to near zero at point of entry.
3-Layer Variance Framework™

Not all waste is the same problem

The 3-Layer Variance Framework is Varensi's core intellectual property. It classifies every waste record into one of three distinct layers — each requiring a completely different operational response.

Layer 01 · Biological Baseline
Natural Yield
Unavoidable prep loss
Natural yield waste from high-quality ingredient preparation. Potato peelings. Vegetable trimmings. Fish bones. Meat trim from butchery. Benchmarkable against industry standards. Never zero, and that's fine.
✓ No action required — monitor and benchmark
Layer 02 · Transient Variance
Operational Noise
Non-recurring anomaly
Non-recurring anomalies identified via confidence scoring. One-off festivals, weather spikes, staffing events, private dining overruns. High confidence the pattern won't repeat. No structural change required.
↗ Monitor only — no structural adjustment needed
Layer 03 · Systemic Attrition
Structural Failure
Recurring process failure
Recurring patterns signalling flawed processes, ordering habits, or menu design. These are fixable — and expensive to ignore every week. Systemic patterns carry both a financial cost and a recoverable margin figure.
⚠ Action required — review par levels and ordering
Regulatory Context

Two mandates. One platform.

UK hospitality operators face two simultaneous compliance obligations in 2026. Varensi addresses both.

Mandatory · October 2026
DEFRA Digital Waste Tracking
Effective: 1 October 2026 · England
All waste producers, carriers, and brokers in England must use the DEFRA DWT system to track waste digitally. Replaces paper Waste Transfer Notes and Waste Consignment Notes. Requires EWC codes, licensed carrier details, consignee facility licences, and independent third-party verifier sign-off.
Varensi: Provides the intelligence layer above DEFRA's basic portal — classification, carbon calculation, pre-submission checklist, and DWT-ready CSV export. The same model as Xero for Making Tax Digital.
Active · April 2026
Simpler Recycling
Effective: April 2026 · England
All businesses in England with 10 or more employees must separate food waste into dedicated streams. Non-compliance risks fines exceeding £5,000 per incident. The requirement affects every hospitality business above the threshold.
Varensi: Waste stream separation is built into the waste type classification system. Every record is categorised at point of entry, providing evidence of separation compliance.
How to Use Varensi

Six steps from entry to compliance

Varensi is designed for operational simplicity. Kitchen staff record waste in seconds. Intelligence, compliance, and carbon reporting happen automatically.

01
Sign in at varensi.com
Authorised users access Varensi at varensi.com. Each account is scoped to your organisation — your data is never visible to other clients. Access is controlled by role: owner, manager, data entry, verifier, or read-only.
02
Record a waste entry
Three ways to record waste: manually via the form (site, waste type, weight, date), by photographing the waste bin and letting the intelligence engine identify and auto-fill the entry, or by speaking a description and having the engine transcribe and classify it.
Photo and voice entry available on mobile
03
Varensi classifies it automatically
Within seconds, the Varensi Intelligence Engine classifies the record into one of the three variance layers — Biological, Transient, or Systemic — with a confidence score and a plain-English reason. CO₂e is calculated instantly using DEFRA emission factors. Everything is written to an immutable audit trail.
04
Review your Intelligence Dashboard
The Intelligence tab shows your recovery potential, 3-Layer breakdown, carbon impact, and all records. The Carbon tab shows your Scope 3 Category 5 totals and DEFRA factor reference. Ask the Intelligence Engine questions about your data at any time.
05
Work through the Comply checklist
The Comply tab contains your DWT readiness score, an interactive 9-item pre-submission checklist, the responsibility matrix, and the October 2026 countdown. Each checklist item saved is timestamped and attributed to the completing user — this is your compliance audit trail.
Checklist state saved to database — visible to all authorised users
06
Export your DWT-ready report
When your checklist is complete and your independent verifier has signed off, export your DWT-ready CSV from the Comply tab. The export maps directly to DEFRA submission fields and includes AI disclosure flags for regulatory transparency.
All exports must be independently verified before DEFRA submission
Frequently Asked Questions

Everything you need to know

Questions from sustainability managers, GMs, CSOs, legal teams, and IT security. If your question isn't here, email hello@mediahubink.com.

Compliance & Regulatory
From 1 October 2026, all waste producers, carriers, and brokers in England must use the DEFRA Digital Waste Tracking system to track waste digitally. It replaces paper Waste Transfer Notes (WTNs) and Waste Consignment Notes (WCNs). Every movement of waste must be recorded with EWC codes, licensed carrier details, consignee facility information, and independent verifier sign-off. Non-compliance risks enforcement notices, financial penalties, and potential prosecution. DEFRA provides a free basic portal — Varensi is the intelligence layer above it, following the same model as Xero for Making Tax Digital.
Not at current pilot stage. Varensi exports a DWT-ready CSV that maps directly to DEFRA submission fields, which you then upload to the DEFRA portal. Direct API integration with DEFRA's Receipt of Waste API is in development and will be live before the October 2026 mandate. All exports must be independently verified before submission — this is a DEFRA requirement, not a Varensi limitation.
Source data entry, EWC code verification, carrier details, final DWT submission, and independent audit remain 100% the client's sole responsibility. Varensi provides intelligence, analysis, and suggested exports only. Varensi accepts no liability for data accuracy or regulatory non-compliance. This is clearly stated in the Responsibility Matrix in the Comply tab. All exports must be independently verified by a qualified verifier before regulatory submission.
From April 2026, all businesses in England with 10 or more employees must separate food waste into dedicated streams. Fines exceed £5,000 for non-compliance. Varensi's waste type classification system categorises every record at point of entry — providing documented evidence of separation compliance across all sites.
AI & Technology
Varensi's intelligence engine is built on enterprise-grade AI. The specific technology is not disclosed for commercial reasons. All AI classifications are explainable (every decision includes a plain-English reason), auditable (every classification is stored in an immutable audit log), and subject to human override (any classification can be overridden with a documented reason). This architecture complies with UK DSIT AI governance requirements and EU AI Act transparency obligations.
Yes. The UK AI governance framework (DSIT 2024–2026) requires AI systems in regulated contexts to be transparent, auditable, human-overseen, and accurate. Varensi complies in full: Transparency — every AI classification includes a one-sentence reason stored in the database. Auditability — an immutable audit log captures every change with before/after snapshots, user attribution, and timestamp. Human oversight — any classification can be overridden by an authorised user, with the override reason mandatory and permanently recorded. Accuracy disclosure — a confidence score (0–100%) is stored and displayed for every classification. DEFRA's guidance that AI-assisted data entry must be disclosed in the submission audit trail is addressed by the human override field and AI disclosure flags in the DWT export.
Disclose Varensi AI as Scope 3, Category 1 (Purchased Goods and Services — Software/SaaS). Annual estimate: approximately 0.06 tCO₂e per year. Method: consumption-based (kWh × UK grid intensity 194 g/kWh, 2026). This follows GHG Protocol Scope 3 Standard. Varensi's Carbon tab provides the AI emissions data required for this disclosure. Not included in the estimate: client-side rendering, data transmission, or verifier device usage.
Data & Security
All data is stored in Supabase EU-West-2 (London) — UK data residency, compliant with UK GDPR. Row Level Security (RLS) is enforced at the database level: your organisation's data is completely isolated from all other clients. Even Varensi's own team cannot access client data without explicit authorisation. The database uses AES-256 encryption at rest and TLS 1.3 in transit. Supabase holds SOC 2 Type II certification.
You own your data. Varensi acts as a data processor under UK GDPR, not a data controller. Your waste records, classification decisions, audit trail, and compliance checklist state are yours. All data is exportable in full at any time via CSV. On contract termination, you receive a complete data export within 30 days. Data is retained for 7 years (HMRC requirement) and deleted on request thereafter.
All client data is exportable in full at any time via CSV — you are never locked in. On contract termination for any reason, clients receive a complete data export within 30 days. The underlying database infrastructure (Supabase) is independent of Varensi and would continue to operate. For enterprise clients, data continuity provisions can be written into the contract.
Product
The 3-Layer Variance Framework™ is Varensi's proprietary waste classification methodology, developed by Mediahubink. It classifies every waste record into one of three layers: Biological Baseline (unavoidable yield loss — no action), Transient Variance (non-recurring anomalies — monitor only), or Systemic Attrition (recurring process failures — action required). The framework is trademark pending under Class 42 (UK IPO). It is the intellectual property of Mediahubink and may not be reproduced without written permission.
If an authorised user disagrees with the AI classification, they can override it manually. The override reason is mandatory — it cannot be left blank. Both the AI's original classification and the human override (with reason, user name, and timestamp) are stored permanently in the audit trail. This dual record is specifically required by DEFRA's guidance on AI-assisted data entry in DWT submissions and satisfies UK AI governance human oversight requirements.
Varensi is architected for multi-site operations from the ground up. There is no technical limit on the number of sites per organisation. The Portfolio tab provides cross-site comparison. Data is isolated per site within your organisation. New sites can be added directly from the Portfolio tab without any code changes. Enterprise plans include dedicated account management for groups with 10+ sites.
Enterprise Brief

For procurement, legal, and C-suite review

Answers to the questions that sustainability directors, CFOs, legal counsel, and IT security teams ask before signing.

What Varensi provides
  • Waste classification using the 3-Layer Variance Framework™
  • CO₂e calculation using DEFRA UK 2024 emission factors
  • DWT pre-submission checklist and compliance readiness scoring
  • DWT-ready CSV export mapped to DEFRA submission fields
  • Immutable audit trail with user attribution and timestamps
  • AI emissions tracking and ESG disclosure guidance
  • Multi-site portfolio management and cross-site comparison
  • Intelligence engine for natural language compliance queries
What remains the client's responsibility
  • Source data accuracy and verification
  • EWC code confirmation with licensed waste carrier
  • Waste carrier licence verification (Environmental Permitting Register)
  • Consignee facility licence confirmation
  • Appointment and management of independent verifier
  • Independent verifier sign-off before DEFRA submission
  • Final DEFRA DWT portal submission
  • AI emissions inclusion in ESG reporting
Infrastructure and security
Data residency: Supabase EU-West-2 (London) · UK GDPR compliant · Encryption: AES-256 at rest, TLS 1.3 in transit · Certification: Supabase SOC 2 Type II, Vercel SOC 2 Type II · Authentication: Supabase Auth with JWT tokens, 8-hour session expiry · Uptime: 99.99% (Vercel SLA) · Isolation: Row Level Security enforced at database level — complete data isolation between organisations
AI governance compliance
UK DSIT framework: Transparency ✓ · Auditability ✓ · Human oversight ✓ · Accuracy disclosure ✓ · EU AI Act (Article 52): Limited risk category, transparency obligations met · DEFRA DWT AI guidance: AI-assisted classification disclosed in audit trail and export · ISO 27001: Target 2027 roadmap — current infrastructure operates on SOC 2 Type II certified providers
Commercial and contractual
Data ownership: Client owns all data at all times · Portability: Full CSV export available on demand · Termination: Complete data export delivered within 30 days · Retention: 7 years (HMRC requirement), deletion on request · Liability: Varensi accepts no liability for data accuracy or regulatory non-compliance — client is sole responsible party for DWT submission · Contact: hello@mediahubink.com for DPA and contract terms

October 2026 is 5 months away.

Every hospitality operator in England with waste obligations needs to be DWT-ready. Varensi runs a free 60-day pilot for qualifying operators — real data, real classification, real compliance progress.

No commitment required for the pilot · hello@mediahubink.com · linkedin.com/in/joashf-perera