Back to Oil & Gas Hub
Downstream

Commercial & market intelligence

One of many downstream use cases. This example focuses on crude acquisition and margin analysis, helping trading and planning teams optimize the spread between feedstock cost and product value.

The information asymmetry problem

Crude acquisition decisions involve balancing quality (assay characteristics), logistics (pipeline access, shipping), pricing (differentials, term vs. spot), and refinery capability (unit constraints, product slate targets). Your commercial team processes this information from multiple sources: brokers, market reports, internal planning, and makes decisions under time pressure.

The traders who excel have decades of pattern recognition. When they retire, that intelligence leaves with them.

$0.50-2/bbl
Margin improvement from optimized crude acquisition
100+
Crude grades available: each with unique economics
$10M+
Annual impact of crude selection optimization for a mid-size refinery

Why current tools fall short

Traditional approaches were not built for the scale and complexity of modern operations.

01

Spreadsheet-based crude economics

Crude netback calculations in Excel don't update with real-time market pricing or refinery yield changes.

02

Delayed market intelligence

By the time broker reports are compiled and analyzed, trading windows have closed.

03

No integrated view

Crude assays, market pricing, logistics costs, and refinery yields live in separate systems. Connecting them is manual work.

04

Historical pattern loss

Market relationships (crude differentials, seasonal patterns, grade substitution economics) are known by experienced traders but not codified.

The Lumina Approach

How Lumina solves it

Three layers of intelligence working together: reasoning agents, proactive detection, and multi-agent deliberation.

Layer 1

Commercial intelligence engine

Agents evaluate crude economics by connecting assay data, current market pricing, logistics costs, and refinery-specific yield predictions. Every analysis runs as SQL against your commercial and operational data.

Every calculation is SQL you can verify. No black box.

Agent Reasoning
SQL-Backed
Agent Reasoning:

Evaluating Q2 Crude Acquisition Options: WCS (Hardisty): $58/bbl, pipeline cost $4.50, yield value $71.20 → Net margin: $8.70/bbl Cold Lake Blend: $61/bbl, pipeline cost $4.50, yield value $72.80 → Net margin: $7.30/bbl Syncrude: $72/bbl, pipeline cost $3.20, yield value $78.40 → Net margin: $3.20/bbl → WCS delivers best margin but requires 95% coker utilization → At current coker rate (88%), blend 75% WCS + 25% Syncrude optimizes net margin: $7.40/bbl

All outputs backed by verifiable SQL you can inspect
The Radar
Scanning
Anomaly Detected:

ALERT: WCS-WTI differential widened to -$22/bbl (30-day avg: -$15) Historical pattern: 3 of last 4 widening events >$20 reverted within 2 weeks Cause analysis: Temporary pipeline apportionment increase (Enbridge Line 3) → Opportunity: Lock in term barrels at current differential. Expected reversion value: $3.50/bbl on 30-day forward

Proactive hypothesis testing, like anomaly clustering
Layer 2

Market anomaly detection

The Radar monitors crude differentials, product crack spreads, and market structure for unusual patterns that signal trading opportunities or risks.

The Radar surfaces issues the operator didn't know to look for. Before they become incidents.

Layer 3

Commercial strategy deliberation

Trading, planning, and risk agents reason together on crude acquisition, hedging strategy, and product slate optimization.

The output is grounded in facts (SQL results), not hallucination. Every recommendation carries a full audit trail.

The Boardroom
Deliberation
Multi-Agent Debate:

Trading Agent: WCS differential at -$22 is an opportunity. Recommend increasing term commitment Planning Agent: Coker turnaround in Q3 limits heavy crude processing for 45 days. Can't increase term above Q2 Risk Agent: Differential volatility is elevated. Recommend hedging 50% of incremental barrels with basis swaps → Consensus: Increase Q2 WCS term by 5,000 bbl/d at current differential. Hedge 50% with basis swap. Reduce Q3 term to match coker outage.

Agents vote, challenge, and produce a synthesized recommendation

The result: intelligence that scales

Lumina addresses the four strategic problems that hold operators back.

Scale without headcount

Evaluate every crude grade, every day, against current market conditions and refinery-specific economics.

Reduce reactive costs

Capture $0.50-2/bbl in margin improvement through optimized crude acquisition and product slate management.

Knowledge retention

Market relationships, grade economics, and trading heuristics are encoded, not dependent on individual traders' experience.

Auditable trust

Every netback calculation, differential analysis, and margin estimate is SQL-verifiable against your commercial and market data.

Agents for this use case

Specialized AI agents that power this workflow.

Lumi Trade

Commercial Analyst

Specializes in crude economics, netback analysis, differential monitoring, and grade substitution evaluation.

Lumi Market

Market Intelligence

Monitors crack spreads, benchmark correlations, seasonal patterns, and market structure changes.

Lumi Core

Data Scientist

Handles market data integration, assay databases, historical pattern analysis, and correlation studies.

This is just one of many use cases

Explore what Lumina can do for your operation

Commercial & market intelligence is one example of how Lumina reasons on operational data. Across Oil & Gas, every domain has use cases where AI agents can add value.

Not ready to commit? Stay up to date as we release new capabilities and industry-specific agents.

View all Oil & Gas use cases