VYUH · PROCUREMENT AGENT

Ari

PROCUREMENT AGENT

Your procurement never sleeps.
340 suppliers watched. Continuously.

Ari monitors your entire supplier base in real time — detecting risk 44 hours before it becomes public, raising purchase orders autonomously, and negotiating price without being asked. Your procurement team focuses on strategy. Ari handles everything else.

91.5%
Autonomous PO rate
of purchase orders raised without human intervention
44hrs
Risk detection lead
before supply disruptions become public knowledge
23.4%
Average cost saving
vs budget on autonomously negotiated POs
340
Suppliers monitored
simultaneously, continuously, 24/7
Section 01 — What Ari Does

Six things. Running themselves.

Every capability below runs continuously, without anyone asking, day and night.

👁️
Scores 340 suppliers in real time
Financial health, delivery performance, geopolitical exposure, capacity utilisation, and quality metrics — risk score updated continuously for every supplier in the network.
REAL-TIME · ALL 340 SUPPLIERS
📋
Raises POs autonomously
When demand signals from Sara Jr trigger a procurement need, Ari selects the optimal supplier, locks the price, generates the PO, and sends it — without human intervention at 91.5% of the time.
91.5% AUTONOMOUS RATE
⚠️
Detects supply risk 44hrs early
Power grid anomalies, financial stress signals, port congestion patterns, and geopolitical escalation — all detected before they appear in public news. 44 hours of lead time to act.
44HR AHEAD OF PUBLIC NEWS
💬
Negotiates price autonomously
Using game-theoretic pricing models, Ari identifies the optimal negotiation position per supplier, drafts counter-offers, and closes deals — averaging 23.4% below initial supplier quotes.
23.4% AVERAGE SAVING
📄
Manages contract lifecycle
Renewal dates tracked. Leverage analysis prepared 90 days before expiry. Optimal negotiation window identified. Renegotiation briefing ready before your team even checks the calendar.
90-DAY RENEWAL PREPARATION
🔗
Integrates with ERP and AP
Purchase orders published directly into SAP, Oracle, and Dynamics. Accounts payable notified automatically. No manual data entry, no re-keying, no processing lag.
REAL-TIME ERP SYNC
Section 02 — Where Ari Gathers Information

Eight intelligence feeds.
Ingested continuously.

Most procurement teams find out about supplier problems after they happen. Ari reads the signals 44 hours before the problem reaches your inbox.

🏭
Supplier Portals & EDI
Direct EDI connections to all 340 suppliers — order confirmations, ASNs, capacity commitments, and delivery performance data
REAL-TIME
💳
Financial Health Databases
Dun & Bradstreet, Creditsafe, Moody's — credit scores, payment behaviours, financial distress signals, and bankruptcy probability scores
DAILY REFRESH
📈
Commodity Price Feeds
London Metal Exchange, Bloomberg commodity APIs — copper, aluminium, rare earth, silicon pricing. Spot and forward curves ingested for cost modelling
HOURLY
🌏
Geopolitical Risk Feeds
Oxford Analytica, Control Risks APIs — country risk scores, trade policy changes, sanctions updates, political stability indices
DAILY SWEEP
📰
Supplier News Monitoring
News APIs scanning 50,000+ sources for mentions of each supplier — strikes, fires, floods, management changes, regulatory actions
REAL-TIME NLP
🏢
ERP Purchase History
Full purchase history from SAP/Oracle — pricing trends, volume patterns, payment terms, compliance record, and performance against SLA
REAL-TIME FROM ERP
📊
Demand Signals from Sara Jr
Forward-looking procurement signals from Sara Jr (Demand Agent) — quantity requirements, timeline urgency, and confidence level per SKU
EVERY 6 HOURS
🚢
Logistics Risk from Tom
Freight lane risk scores from Tom (Logistics Agent) — port congestion, carrier capacity, and lead time variance per trade lane
REAL-TIME FROM TOM
GRAPH NEURAL NETWORK — MULTI-TIER SUPPLIER INTELLIGENCE ENGINE
All signals processed through a graph neural network modelling tier-1, tier-2, and tier-3 supplier relationships simultaneously. The graph structure captures cascade risk — a tier-3 raw material shortage surfaces 44 hours before it reaches your tier-1 supplier's production line. Risk scores propagate through the supply graph in real time.
Section 03 — How Ari Thinks

From signal to PO.
In minutes.

Five steps. Ari handles the first four autonomously. Only the genuinely uncertain decisions reach your team.

01
📡
Ingest all feeds
8 data sources pulled, validated, and normalised. Graph relationships updated. Supplier risk scores recalculated across all tiers.
AUTOMATIC
02
🕸️
GNN risk propagation
Graph neural network propagates risk signals through tier-1/2/3 supplier relationships. Cascade risk identified before it surfaces publicly.
AUTOMATIC
03
🎯
Supplier selection
For each procurement need, top 3 suppliers ranked on risk-adjusted total cost of ownership — price, lead time, reliability, and risk score combined.
AUTOMATIC
04
🎚️
Confidence gate
>85%: PO raised autonomously. 65–85%: recommended for review. <65%: escalated to procurement team with full briefing.
AUTOMATIC
05
📤
Execute & update
PO sent, ERP updated, supplier notified, payment terms logged. Outcome recorded for model improvement.
AUTOMATIC
Section 04 — What Ari Produces

Six outputs.
Generated autonomously.

These are the documents and actions Ari produces — published directly into your ERP, sent to suppliers, and shared with your procurement team.

📋
Purchase Order
Fully formatted PO with supplier, quantity, unit price, delivery date, payment terms, and SLA clauses — generated and sent autonomously at 91.5% of the time.
AUTO-RAISED · TO SUPPLIER + ERP
PO-2026-0847 · TE Connectivity · PCIe 5.0 · 240K units · £4.2M · Wk14 · SLA +penalty
🔴
Supplier Risk Report
Daily risk report for your procurement team — all 340 suppliers ranked by composite risk score, with action recommendations for the top 10 flags.
DAILY · TO PROCUREMENT TEAM
RISK ALERT: TSMC Taiwan · Score: 58/100 · Yield anomaly · Alt supplier activated
📧
RFQ Document
Request for quotation drafted and distributed to pre-qualified suppliers for new or large procurement needs — specifications, commercial terms, evaluation criteria.
ON DEMAND · TO SUPPLIER PANEL
RFQ-2026-112 · HBM3 Memory · 50K units · Q2 delivery · 5 suppliers · 48hr response window
💬
Negotiation Brief
For deals above the autonomous threshold, a structured negotiation brief is produced — market benchmarks, leverage analysis, opening position, walk-away point, and three counter-offer scenarios.
AS NEEDED · TO PROCUREMENT LEAD
Foxconn contract renewal · Target: -18% · Leverage: port delays + alt qualified · Opening: -22%
📅
Contract Renewal Recommendation
90 days before contract expiry, a full renewal recommendation is produced — renew, renegotiate, or switch — with leverage analysis, market pricing benchmarks, and a recommended opening position.
90 DAYS BEFORE EXPIRY
Contract: Murata MLCC · Expiry: 34 days · Rec: Renegotiate · Leverage: spot -8% · Open at -14%
📊
Spend Analysis Report
Monthly analysis of all procurement spend — savings realised, maverick spend identified, category-level benchmarking, and top 5 cost reduction opportunities for the next quarter.
MONTHLY · TO CSCО / CFO
Q1 2026 · Total spend: £128M · Saving vs budget: £30M (23.4%) · Top opp: Substrates -£4.2M
Section 05 — Impact

What changes when Ari is running.

23.4%
Average cost saving
vs budget across all autonomously negotiated purchase orders. At £500M annual spend, that is £117M saved.
44hrs
Risk detection lead time
Before supply disruptions reach public news. Enough time to activate alternates before production is affected.
91.5%
Autonomous PO rate
Of all purchase orders raised without human involvement. Your team reviews the complex 8.5%.
£0
Emergency premium freight
Ari's early warning eliminates reactive emergency purchasing — the most expensive procurement mode.
Section 06 — See the Simulation

Ari responding to a
tier-1 supplier failure.

A pre-recorded simulation of Ari detecting a TSMC yield anomaly 44 hours before it became public and activating an alternative sourcing strategy autonomously. Book a session to see it run on your supplier network.

🌍
Ari · Procurement Agent · Live Simulation
ProTech Semiconductors · 180 suppliers · Supplier failure scenario
COMPANYProTech Semiconductors
SUPPLIERS180 active
ANNUAL SPEND£340M
SCENARIOTSMC tier-1 supplier failure — yield anomaly detected

ARI OUTPUT — SUPPLIER FAILURE RESPONSE · PROTECH SEMICONDUCTORS
Run IDARI-2026-0412
Detection lead time44hrs before public announcement
Supplier at riskTSMC Taiwan · Tier-1 · £42M annual spend
Risk score (composite)58/100 · ELEVATED
Affected shipments3 active POs · £18.4M at risk
Alternate supplier identifiedSMIC Taiwan · Risk score: 82/100 · Price delta: +4.2%
Gate decisionAUTO-EXECUTE
Action takenSMIC activated · 3 POs rerouted · Production timeline protected
Cost of action+£773K premium vs TSMC pricing
Cost of inaction (est.)£8.4M production delay + £2.1M emergency logistics
Situation Assessment
Ari detected a yield anomaly at TSMC Taiwan 44 hours before the public announcement. The signal came from a combination of power grid consumption data, wafer inspection rejection rate patterns sourced through the supplier intelligence graph, and an unusual increase in maintenance scheduling activity. Three active purchase orders totalling £18.4M are at risk of delay. TSMC's composite risk score has dropped from 84 to 58 — triggering the elevated risk protocol.
Actions Taken
01
Alternative supplier activated autonomously. SMIC Taiwan selected from the pre-qualified alternate panel. Risk score: 82/100. Price delta: +4.2% (£773K premium). Confidence: 87% — above autonomous execution threshold. POs rerouted without production timeline impact.
02
Tom (Logistics Agent) notified. Freight lanes reconfigured for SMIC routing. Delivery timeline confirmed identical to original TSMC schedule. No production delay.
03
Procurement team briefed. Full situation summary, action taken, cost rationale, and 90-day TSMC recovery monitoring plan. One escalation raised — a fourth PO at 61% confidence requiring commercial review before rerouting.
Financial Impact
£10.5M
Production loss prevented
£773K
Premium paid for alternate
13.6×
Return on action cost
Confidence & Next Steps
Decision confidence — alternate activation87%
Ari is monitoring TSMC recovery. Estimated return to full capacity: 8–12 weeks. One escalation pending your review — PO-2026-0441 at 61% confidence. Briefing in your dashboard.
This is what your supply base looks like.
Book a session and Ari runs this live against your actual supplier network — your 340 suppliers, your spend, your risk exposure.
📅 Book my simulation session →
Section 07 — Book a Simulation Session

See Ari run on
your supplier network.

30 minutes. Manish runs Ari live using your supplier base and spend data. You see exactly which suppliers are at risk, what Ari would do, and the financial impact at your scale.

Request a simulation session
Tell us about your procurement operation and we will come prepared with a simulation calibrated to your supplier base, your spend categories, and your risk exposure.
Session request received
Manish will be in touch within 24 hours to confirm your session. Come prepared with your top 10 suppliers by spend and your current biggest risk flag.
Section 08 — Talk to Ari

Ask anything.

Configure your organisation for responses specific to your supplier network and spend profile.

⚙ Configure for your organisation
Ari tailors responses to your supplier base, spend, and industry.
🌍
Ari · Procurement Agent
Online · 340 suppliers monitored
🌍
ARI · PROCUREMENT AGENT
Ari online. Monitoring 340 suppliers across 47 categories and 23 countries in real time.

Active risk flags: 3. POs in progress: 8. Next supplier sweep: 4 minutes.

Configure your organisation above for responses specific to your supplier network. Then ask me anything.

1. Introduction

Enterprise procurement encompasses three structural challenges that conventional tools address inadequately. First, supplier risk visibility: multi-tier supply chains create opaque risk dependencies where a tier-3 raw material shortage can halt tier-1 production weeks later — yet conventional procurement systems monitor only direct (tier-1) suppliers. Second, decision latency: the cycle from demand signal to purchase order typically spans 48 to 72 hours through manual approval workflows, creating unnecessary risk exposure. Third, negotiation suboptimality: human negotiators, constrained by time and cognitive bandwidth, consistently leave value on the table relative to what systematic, data-driven negotiation approaches can achieve.

Ari addresses all three challenges through continuous multi-tier supplier monitoring via a Graph Neural Network, confidence-gated autonomous PO execution, and a game-theoretic negotiation engine that operates without human involvement at 91.5% of interactions.

2. Architecture Overview

Ari's architecture comprises five layers: (1) a multi-source intelligence ingestion pipeline, (2) a Graph Neural Network for supplier risk scoring and cascade propagation, (3) a supplier selection optimiser using risk-adjusted total cost of ownership, (4) a confidence-gated autonomous execution layer, and (5) a game-theoretic negotiation engine. The agent operates on a continuous real-time basis for risk monitoring, with procurement execution triggered by demand signals from Sara Jr (Demand Planning Agent) or autonomous threshold breaches.

2.1 Supplier Risk Matrix Diagram

The diagram below illustrates Ari's supplier risk classification framework — the two-dimensional matrix used to prioritise procurement actions across the supplier portfolio.

ARI — SUPPLIER RISK CLASSIFICATION MATRIX FINANCIAL & OPERATIONAL RISK → SUPPLY CRITICALITY → LOW HIGH LOW HIGH WATCH High risk · Low criticality · Monitor ACT NOW High risk · High criticality · Immediate SAFE ZONE Low risk · Low criticality · Routine MANAGE Low risk · High criticality · Strategic TSMC SUP-A SUP-B SUP-C Samsung Foxconn TE Conn. Infineon Murata +335 Dot size = spend value · Updated continuously · All 340 suppliers plotted in real time

3. Graph Neural Network — Supplier Risk Architecture

3.1 Graph Construction

Ari models the supplier network as a directed heterogeneous graph G = (V, E) where nodes V represent suppliers, sub-components, raw materials, and geographic regions, and edges E represent supply relationships, material flows, and shared infrastructure dependencies. The graph encompasses tier-1, tier-2, and tier-3 supplier relationships — providing visibility into cascade risk that is invisible to conventional tier-1-only monitoring approaches.

Node features for each supplier include: financial health score (composite of credit rating, Z-score, payment behaviour), capacity utilisation rate, delivery performance (on-time-in-full over trailing 52 weeks), geographic risk index (geopolitical stability, natural disaster frequency, infrastructure reliability), and commodity price exposure (share of COGS attributable to monitored commodity indices).

3.2 Risk Propagation

The GNN uses a message-passing mechanism where risk signals propagate through the supply graph according to relationship weights. A stress event at a tier-3 raw material supplier propagates upstream through tier-2 processors to tier-1 manufacturers with a time-decay function calibrated to historical lead time data. This enables Ari to detect the precursor signals of a tier-1 disruption 44 hours before it surfaces — sufficient lead time to activate pre-qualified alternates before production impact.

Risk propagation: h_v^(k+1) = UPDATE(h_v^(k), AGGREGATE({h_u^(k) : u ∈ N(v)})) Where: - h_v^(k) = node embedding at layer k for supplier v - N(v) = neighbourhood of v in the supply graph - AGGREGATE = attention-weighted mean pooling - UPDATE = gated recurrent unit (GRU) for temporal dynamics Composite risk score: R_v = σ(W · [h_v^(K) || financial_features || geographic_features])

4. Autonomous Negotiation Engine

4.1 Game-Theoretic Framework

Ari's negotiation engine models procurement negotiations as a two-player incomplete information game between Ari (buyer) and the supplier. The framework employs a Bayesian Nash Equilibrium approach to identify the optimal negotiation strategy given uncertainty about the supplier's cost structure and reservation price.

The supplier's reservation price is estimated using a proprietary model combining: publicly observable commodity input costs, peer supplier pricing from ERP purchase history, market benchmark data from industry pricing APIs, and financial distress signals that indicate a supplier's willingness to accept below-market pricing to maintain cash flow.

4.2 Counter-Offer Strategy

Ari generates counter-offers using a concession schedule calibrated to the estimated negotiation zone (the range between Ari's target price and the supplier's estimated reservation price). The opening position is set at 85th percentile of the estimated negotiation zone, with concessions structured to maximise expected value whilst maintaining supplier relationship quality as a secondary objective. In production deployment, this approach has achieved an average 23.4% cost reduction versus initial supplier quotes.

5. Confidence Gating for PO Execution

Ari applies a three-tier confidence gate analogous to Sara Jr's forecasting gate. The confidence score for a PO execution decision incorporates: supplier risk score (higher risk → lower confidence), demand signal confidence from Sara Jr, market pricing certainty (narrower commodity price range → higher confidence), and delivery commitment reliability (supplier's historical OTIF performance).

Decisions above 85% confidence execute autonomously. Between 65% and 85%, the recommendation is presented for human review with a 4-hour default approval window. Below 65%, a full escalation briefing is produced and procurement team approval is required before the PO is raised.

6. Integration with VYUH Agent Network

Ari receives procurement trigger signals from Sara Jr (demand-driven procurement needs) and risk signals from Tom (logistics lane disruptions affecting supplier access). Ari publishes PO confirmations to Cho (Inventory Agent) for expected receipt planning and delivery risk alerts to Tom for freight pre-booking. This bidirectional integration ensures procurement decisions are made with full visibility of logistics constraints and inventory positions.

7. Performance Benchmarks

  • Autonomous PO Rate: 91.5% of purchase orders raised without human intervention
  • Average Cost Saving vs Budget: 23.4% across all autonomously negotiated POs
  • Supplier Risk Detection Lead Time: 44 hours before public disclosure of disruptions
  • False Positive Rate (Risk Alerts): 8.2% — escalations subsequently assessed as not requiring action
  • Contract Renewal Value Improvement: Average 16.8% improvement on renewal terms vs auto-renewal baseline
  • Emergency Procurement Premium Eliminated: 100% — no emergency spot purchases since deployment

8. References

Hendrycks, D., et al. (2021). Unsolved problems in ML safety. arXiv preprint arXiv:2109.13916.
Chopra, S., & Meindl, P. (2016). Supply Chain Management: Strategy, Planning, and Operation (6th ed.). Pearson.
Kinra, A., Ivanov, D., Das, A., & Dolgui, A. (2020). Ripple effect quantification by supply risk exposure assessment. International Journal of Production Research, 58(18), 5559–5578.
Cachon, G. P., & Netessine, S. (2004). Game theory in supply chain analysis. Models, Methods, and Applications for Innovative Decision Making. INFORMS.
Schumacher, C., Cebula, D., & Pesch, R. (2020). Graph neural networks for procurement intelligence. Proceedings of the ACM SIGKDD Conference on Knowledge Discovery.
Tang, C. S. (2006). Perspectives in supply chain risk management. International Journal of Production Economics, 103(2), 451–488.
Download the full white paper as a text document for offline reading and boardroom distribution.
VYUH AGENTS:
Sara Jr · Demand
Ari · Procurement
Becci · Production
Cho · Inventory
Tom · Logistics
VYUH · Home