📡
🏢
CH
Launchpad
PETS.L — FY26
📡 Radar 3
Library
Coverage
Architecture

Good morning, Charlie

What are you working on today?

Multi-Lens ▼
📎 UploadSources 5+ Notes⚙ 4 LensesPlannedYOLO
Templates See all (42)
All
▦ Model
⚖ Verify
⚠ Red Team
📈 Valuation
⚡ Trigger

M Simple Model Build

Multi-model extraction verification.

Model4 Lenses~15 min

K Deep KPI Extraction

Dual-model with divergence flags.

Model2 Lenses~10 min

S Metric Standardisation

Peer metrics cross-checked via FMP.

Model3 Lenses~20 min

A Accounting Screen

M-Score, forensic checks, BS+CF verified.

Model+Note4 Lenses~15 min

R Anti-Thesis Build

3 models construct the bear case independently.

Note3 Models~10 min

V Quick Valuation

Live EV/EBITDA, P/E, FCF yield, peer comps.

ModelLens 4~5 min

C Contradiction Scan

CEO statements cross-referenced over 4 quarters.

NoteLens 3~8 min

P Peer Disclosure Gap

If peers flag a risk and you don't — silence is signal.

Research2 Lenses~12 min

📡 Radar Setup

Automated multi-lens monitoring. Cron alerts on divergence.

Trigger4 LensesAuto
✪ Multi-Lens Synthesis 4/4
Stabilisation confirmed, 3 structural concerns. FY26 uPBT £92.8m (-30%). Vet ~90% of profit — concentration risk. Cross-Statement: negative tangible equity (-£8.6m). Contradiction lens flagged CMA + net debt. Neutral at 192p (6.05x EV/EBITDA, 17.8% FCF yield).
Revenue
£1,469.6m
-0.8%
Underlying PBT
£92.8m
-30.2%
FCF
£147m
17.8% yield
Tangible Equity
-£8.6m
GW: £960m
🔎 Extraction
⚖ Cross-Stmt
⚠ Contradiction
📈 Market
MetricGPT-5.5Claude OpusStatusFMP Verified
Group uPBTc£92mc£92mMatch£92.8m
Retail uPBTc£30mc£30mMatch~£30m
Vet Group PBTc£83mc£83mMatch~£83m
Net Debtc£20mc£20mPartial£357m inc leases
DPS Change-28%-35%Diverge-43.1%
CheckValueFinding
Goodwill/Equity98.7%Tangible equity negative. Impairment = £96m write-down
Gross Margin (3yr)46.8→46.9→45.7%120bp structural compression
OpCF/PBT2.05xStrong cash conversion
FCF/DPS4.3xDividend well covered
Net Debt Defs£20m vs £357m£337m lease liabilities excluded
ClaimCounter-EvidenceSeverity
"Net debt c£20m"Total debt £397m. 95% excludedMisleading
"Comfortable with £99m"FY26 guide was £115-125m, actual £92.8mCredibility
"No CMA impact"CMA text not analysed independentlyUnverified
"£20m savings done"Central costs rose £15.8→£21mUnreconciled
EV/EBITDA
6.05x
P/E
~13x
FCF Yield
17.8%
Div Yield
3.9%
Analyst Consensus (11)
5 Buy3 Hold3 Sell222p (+15.6%)
BrokerRatingTarget
JefferiesBuy265p
CanaccordBuy245p
BerenbergHold

⚙ Key Flags 4

Negative Tangible Equity
-£8.6m. GW £960m = 98.7%
Lens 2
Gross Margin -120bps
Structural. ~£17.6m impact
Lens 2
Vet Concentration 90%
Single-segment dependency
Lens 4
FCF £147m
17.8% yield. DPS 4.3x covered
Lens 2

📋 Monitoring

1Gross margin recovery
2Central cost breakdown
3Goodwill impairment (IAS 36)
4Retail LFL magnitude

⚖ Lens Confidence

Extraction5/6
Cross-Stmt2 Crit
Contradiction4 Flags
MarketLive

💬 Sources

📄 FY26 Pre-Close
📈 FMP API (P&L, BS, CF)
📊 MarketScreener
📄 FY26 Prelim Results

📡 Radar

Automated multi-lens monitoring. Cron jobs run every 6 hours across your coverage universe. Only surfaces findings that pass the materiality threshold.

PETS.L — Negative Tangible Equity Detected
Cross-statement scan found goodwill of £960m against equity of £973m. Tangible book value is -£8.6m. This was NOT flagged in the pre-close statement or any sell-side note.
CriticalLens 2: Cross-StatementAuto-verified via FMP
2 hours ago
IWG.L — Beneish M-Score Breach (DSRI 1.77)
Trade receivables jumped from $368m to $651m while revenue was flat. M-Score crossed manipulation threshold. Normalised in FY25 but pattern warrants monitoring.
CriticalLens 2: Forensic ScreenCron: Daily 07:00
6 hours ago
TSCO.L — Management Narrative Shift Detected
Contradiction lens found CEO's Q3 language on "accelerating online growth" contradicts Q2's "moderating digital mix." Sentiment shift: +0.4 to -0.2 on digital strategy confidence.
WarningLens 3: ContradictionCron: Post-earnings
1 day ago
BME.L — Peer Disclosure Gap: Supply Chain Risk
3 of 4 discount retail peers (COST, DG, DLTR) flagged Red Sea supply chain disruption in Q3 calls. B&M did not mention it. Silence may be signal.
WarningLens 3: Peer Cross-ReferenceCron: Weekly
3 days ago
CVS.L — Valuation Approaching Historical Low
EV/EBITDA at 10.2x vs 5-year average of 14.8x. FCF yield at 8.4% vs average 5.1%. Analyst consensus shifted: 2 upgrades in 7 days.
OpportunityLens 4: Market ContextCron: Daily
5 days ago

Architecture Comparison

■ Primer (Current)

ModelsSingle model at a time
AvailableGPT-5.x + Claude (user selects)
Agent"Monolith" — single agent
DataVisible Alpha (~$50K/yr)
VerificationNone — single model output
Cross-StatementNot implemented
ContradictionNot implemented
ValuationNot in reports
Proactive AlertsTriggers (user-configured)
Report AuthNone (public S3)
BackendHetzner + Render
Tools47+ (monolith arch)

✪ MultiLens (Proposed)

Models2+ models simultaneously
AvailableGPT-5.x + Claude + DeepSeek
AgentSpecialised stateless agents
DataFMP API (~$30/mo)
VerificationDivergence detection engine
Cross-StatementAuto P&L→BS→CF checks
ContradictionRed team agent (Lens 3)
ValuationEvery report, auto-pulled
Proactive AlertsRadar — cron + materiality
Report AuthSigned URLs (Supabase)
BackendVercel + Supabase
ToolsSpecialised per-lens

Multi-Lens Pipeline

Filing Uploaded GPT-5.5 Extract Claude Opus Extract Divergence Check FMP: BS + CF Pull Cross-Statement Verify Contradiction Scan Market Context Synthesis Engine

Each node is a stateless function. Failures are isolated. The Synthesis Engine combines all lens outputs into a single narrative with flagged divergences. Self-improving: every divergence becomes training data for model fine-tuning.

Radar Pipeline (Automated)

Cron (6hr/daily/weekly) Coverage Universe Multi-Lens Scan Materiality Filter Radar Feed Email / Slack Alert

Unlike Primer's Triggers (user-prompted, single model), Radar proactively scans with multi-lens verification. Only surfaces findings that exceed materiality thresholds — no noise. The M-Score breach on IWG would have been caught automatically.

Tech Stack

LayerPrimerMultiLensAdvantage
FrontendNext.js / RenderNext.js / VercelFaster edge deployment
BackendCustom / HetznerVercel Functions + SupabaseServerless, no infra mgmt
DatabaseUnknown (Supabase ref)Supabase (Postgres)Integrated auth+storage+DB
Financial DataVisible Alpha ($50K/yr)FMP ($30/mo)1,666x cheaper
ConsensusVisible AlphaFMP + MarketScreenerComparable quality, fraction of cost
ModelsOpenAI + Anthropic (1 at a time)OpenAI + Anthropic + DeepSeek (parallel)Verification via divergence
OrchestrationMonolith agentSpecialised stateless agentsTestable, isolated failures
LLM OpsPostHog + SentryHelicone + PostHog + SentryPrompt versioning + cost tracking
ReportsS3 (no auth)Supabase Storage (signed URLs)Authenticated by default
MonitoringTriggers (manual)Radar (automated cron + materiality)Proactive, not reactive

Library

Recent
Model

PETS.L — FY26

4-lens verified. 2 critical flags.

Today⚖ 4 lenses
Note

IWG — Forensic

M-Score breach. DSRI flagged.

Yesterday⚖ 3 lenses
Model

TSCO.L — Q3

Cross-statement clean.

2d ago⚖ 4 lenses
Research

UK Retail Peers

PETS, TSCO, MKS, SBRY comps.

3d ago
Note

CMA Vet Report

Independent regulatory analysis.

1w ago
Model

CVS.L — Annual

Vet sector peer. DCF + multiples.

1w ago

Coverage

CompanyTickerSectorPriceEV/EBITDARadarFlags
Pets at HomePETS.LSpecialty Retail192p6.05x📡 Active2 Crit
IWG plcIWG.LReal Estate184p8.2x📡 Active1 Crit
TescoTSCO.LFood Retail372p7.1x📡 Active1 Warn
CVS GroupCVS.LVeterinary1,045p12.4x📡 ActiveClean
B&M EuropeanBME.LDiscount Retail385p7.1x📡 Active1 Warn

New trigger ×

DailyWeeklyMonthly
Multi-LensSingle Model

⚠ Multi-lens triggers cross-check across 2 models before alerting. Reduces false positives by ~60%.