We help operations leaders decide what to make, what to buy, what to hold, and what to promise.
Portolane is a domain-led supply chain and operations consulting firm. We turn planning, manufacturing, procurement, logistics and fulfilment decisions into better ones — built by practitioners, not a pyramid of analysts. That judgement is encoded into the Portolane Decision Intelligence Platform, where AI agents apply it continuously.
41.87°N, 87.62°W — plotted from what is actually known
Domain comes first. Technology amplifies it.
Most analytics firms sell the technology and hope the domain follows. We start with practitioners who have designed, run and rebuilt supply chain and operations processes in real operating environments — then apply trusted data, analytics, AI and intelligent agents to make those decisions faster, better and more consistent.
Technology should quietly enable better decisions — not be the headline.
More about our approach →Deep domain expertise
→ business decisions
Trusted data
→ analytics → artificial intelligence
Intelligent applications
→ AI agents → business outcomes
Analytics applied across the full operating chain
From demand planning to reverse logistics, we bring reporting dashboards, forecasting and prescriptive analytics to the processes where operational decisions actually get made.
What we know, and how we scale it
Every capability we build has two halves. The judgement comes from twenty-two years of running these processes. The technology is how that judgement reaches every decision, not just the ones someone had time for.
Demand forms across dealer, regional and vertical tiers, with make-to-stock and make-to-order in the same network. Consensus needs dual-cycle governance and traceability of what changed between cycles.
Hierarchical forecasting with reconciliation across tiers, ML demand sensing on order-pipeline signals, and an agent that flags divergence before the cycle closes.
Not all stock serves the same purpose. Some is ageing and needs liquidating, some protects an unreleased commitment, some is the only buffer against an unreliable supplier.
Probabilistic multi-echelon inventory optimisation with service-level simulation, treating ageing and liquidation rules as constraints rather than discovering them late.
A date given to a customer is only real if it was computed against actual capacity and material availability — and only some commitments genuinely matter.
Constraint-based supply optimisation, lead-time variance modelling, and an agent that re-promises automatically when the supply plan moves.
A senior planner runs a specific diagnostic sequence, and knows which exceptions to look at first out of four hundred.
LLM-driven root-cause traversal across the planning graph, with exceptions ranked by revenue and commitment at risk, arriving with the analysis already done.
Only certain data breaks actually corrupt a planning run. Knowing which ones is the difference between noise and a gate.
Anomaly detection across the finished-goods and variant hierarchy, with an agent gating the planning run and holding a full audit trail.
Only some physical events should change a plan. Knowing which is a domain question, not a sensor question.
IoT ingestion and event detection for plant telemetry and track-and-trace, feeding movement and line events directly into planning.
Analytics as craft, not decoration
Senior by design
Engagements are led by practitioners with hands-on supply chain and operations experience — not a pyramid of analysts.
Decisions, not dashboards
We scope around what must improve — what to make, buy, hold and promise — then work back to the data.
Analytics as craft
Forecasting, simulation and optimisation are core capability, applied only where they change a decision.
We stay to the outcome
We carry the work through design, build enablement and adoption — not to a recommendation and no further.
Manufacturing depth, applied broadly
Manufacturing is where our depth is greatest — it is not the boundary. We support operations analytics and performance tracking across the following sectors.
"A portolan is a navigational chart — used to plot a dependable course from what is actually known." That is the discipline we bring to analytics: trusted data first, insight second, action third.
Let's plot a course from your operating data to better decisions.
Whether you are just building trust in your reporting or ready to scale AI agents across routine decisions, we meet you where you are.
Contact Us →