By Ami Daniel, Co-Founder and CEO, Windward
The Shifting Sands…
Geopolitical risks seemingly spring out of nowhere, with few expecting the ragtag Houthi militia to shut down major global waterways for a long period, for example. Regulatory shifts, such as OFAC’s latest advisory and the White House’s changing stance on the Russia-Ukraine war, add to the feeling of vertigo. Tariffs go up and down in a day’s notice on social media.
All three of those factors, along with legacy technology, lead to operational disruption. Manual processes simply cannot cope with the scale and speed of modern trade and geopolitics. And fragmented, unstructured data leads to blind spots and delays.
Unstructured data is often misunderstood. To simplify, every enterprise and organization on Earth has a gold mine of data hiding in papers: contracts, agreements, rates, bills of lading, and many more document types.
The drivers in the maritime and supply/logistics space are clear:
- Geopolitical complexity is rising
- Adversarial tactics are more frequent and adaptive
- Competition demands faster, better decisions
- Organizations have valuable data they can’t yet use
Artificial intelligence (AI) is clearly the answer, but…how overwhelmed are you feeling by the pace of AI and Generative AI (gen AI) evolution these days? In the time it took me to write that question, countless new AI apps were probably launched.
What good is all of this shiny new AI technology if your organization lacks the strategy to adopt it effectively?
If you are feeling overwhelmed and reactive, you are not alone. Many people and organizations feel the same. Hopefully, this blog post can help outline a high-level strategy.
Vertical AI is the Foundation Piece
Bradford Cross, founding partner at DCVC Management, sums it up well: AI that delivers core value must be full-stack, subject-matter driven, and embedded in workflows.
That’s vertical AI, a subject I’ve been passionate about for a long time. Generic tools are unable to help a trade compliance officer analyze a ship’s ownership layers. Or a national agency fuse vessel patterns with crew manifests and suspicious behavior.
Vertical AI thrives when:
- The data is unstructured (bills of lading, crew lists, free-text reports)
- The stakes are mission-critical
- The users need speed, precision, and context – not mere prompts
The Time is Now for Gen AI Deployment
The inflection point for gen AI is 2024-2025, according to McKinsey and Boston Consulting Group. Start using it, or get left behind. Menlo Ventures’ 2024 report on generative AI in the enterprise states that 40% of generative AI spending in 2024 came from permanent budgets. This indicates a shift from experimentation, to execution and embedding AI into core business strategies.
This includes crew rosters, inspection logs, customs filings, vessel behaviors, and chat logs – all based on unstructured data, and all high-signal if used correctly.
Structure the Unstructured with Gen AI
Digitizing entire databases has historically been too complicated and costly for organizations, but now with Gen AI, it’s much easier and quicker.
The first step towards automation and efficiency should be using Gen AI agents to structure your data. Windward’s vertical Gen AI agents can structure customer data and turn documents, vessel behavior, and open-source intelligence into contextual insights.
Why the technology exceeds expectations:
- It extracts insights from unstructured sources, such as cargo manifests, port reports, and customs filings
- It reduces manual research and interpretation time by 80-90%
- It produces auditable briefings on everything from vessel behavior to trade risk – on-demand
Step two is choosing one workflow to automate, in order to proceed gradually. And then successfully adding others.
Automating a complex workflow, such as vessel screening or risk assessments, results in actionable insights and an immediate ROI – by streamlining decision-making processes. This integration reduces the reliance on manual labor, minimizes human error, and accelerates operations, leading to increased productivity and efficiency.
The third step, after some process agents are in place, is building the “orchestrator” agent that will automate the full workflow: pre-fixture, revenue leakage, pre-transaction, portfolio risk management, etc.
The Key to Unlocking AI: The Forward-Deployed Engineer
Windward has shifted how we operate, learning from Palantir’s extremely successful model. We’ve moved from dashboards and alerts, to deploying real Gen AI agents, tuned for maritime intelligence, and embedded directly into government and enterprise operations.
It starts with domain-driven AI, purpose-built for maritime and national operations. And it scales through deep partnerships, not just software licenses.
How does the partnership work?
Most organizations have their own IT department and we are not looking to replace that. What they lack are product muscles – personnel who know how to ship tools that work.
That’s where the Forward-Deployed Engineer (FDE) model comes in.
Pioneered and battle-tested by Palantir, the FDE model has proven to be one of the most effective ways to embed AI into operational workflows. FDEs don’t just write code. They live the mission, sit with users, iterate fast, and build AI that delivers tangible value.
Windward is now bringing this proven model to our markets.
Our FDEs embed onsite to:
- Connect to local data (often siloed or sensitive)
- Co-design custom gen AI agents with analysts and operators
- Integrate with internal systems – compliance, port security, trade ops
- Deliver usable tools in weeks, not quarters
This isn’t consulting. It’s accelerated product engineering with AI that’s aligned to real-world complexity.
Windward’s Model: Partnership at the Core
We’re partnering with government agencies, including Singapore’s Defence Science and Technology Agency (DSTA); digital ministries, and commercial leaders across multiple regions to localize Windward’s AI for regional governance and infrastructure.
This includes embedding FDEs to deliver working tools fast – with the same intensity and discipline that made Palantir a global success. This “AI capability transfer” approach will enable us to align with each partner’s operational goals, data sensitivity, and risk tolerance. An approach that is built around mission relevance and measurable outcomes.
From Pilot Projects to Production in Mere Weeks
Unlike some of the new AI entrants who seem to have suddenly appeared on the scene, Windward has a proven maritime track record, with a combined 200+ years of experience from naval commanders, maritime data scientists, and maritime trade and shipping leaders.
That experience enables me to say the following with confidence: combining vertical AI, gen AI, and embedded teams is the only way to make your unstructured data usable, governable, and operational.
You have your data and a sense of urgency. Windward can equip you with the rest, end-to-end. If you’re ready to scale beyond pilot projects and bring gen AI into the heart of your workflows in mere weeks – let us help.
This is the time to go from concept to capability!
See Windward’s Tech in Action
