foryn

Selected Work

Selected Work

Applied AI strategy, workflow design, and decision-support concepts developed through Foryn.

These examples show how Foryn thinks about AI inside real work. Each item focuses on the same practical question: what has to happen between AI output and a decision people are willing to stand behind?

Logistics, supply chain, transportation planning, opportunity analysis

Logistics Decision Support

Problem

Logistics and supply chain teams make decisions from incomplete, time-sensitive, and constraint-heavy information. Requirements are spread across documents, operational assumptions, eligibility rules, handoffs, and source materials. The risk is not only missing information. It is making the wrong next move with confidence.

Foryn contribution

Foryn developed bid and opportunity analysis workflows for logistics and transportation contexts, including federal logistics research. The work focused on turning complex requirements, operational constraints, and source documents into clearer decisions and better next actions.

Artifact/system type

AI-assisted opportunity analysis, requirement extraction, bid/no-bid support, operational constraint review, and decision-record design.

Why it matters

In logistics, AI is useful only if it improves decisions without weakening accountability. A system has to help people see the right constraints, risks, missing inputs, and handoffs before they commit resources or make operational promises.

Status: Internal concept and proposal-development work based on public-sector logistics research. No government endorsement, award, or deployment is implied.

Import operations, customs classification, HTS review, broker oversight, duty accuracy

Trade Compliance Classification

Problem

Import teams work across product descriptions, vendor records, tariff schedules, customs broker communications, duty exposure, and documentation requirements. AI can help organize and compare that information, but trade classification cannot be treated as a silent model decision.

Foryn contribution

Foryn explored AI-assisted classification workflows for customs and trade-compliance contexts. The work focused on product-data intake, HTS candidate review, evidence capture, broker exception review, uncertainty surfacing, and human approval before a classification is used operationally.

Artifact/system type

AI-assisted classification review, broker oversight workflow, evidence trail, exception handling, and approval record.

Why it matters

Classification affects duty, admissibility, penalties, broker instructions, and official records. A useful system should not simply return a code. It should show the basis for the suggestion, identify missing information, flag uncertainty, preserve reviewer judgment, and create a record of what was accepted or changed.

Status: Internal concept and applied workflow research. No customs authority, government endorsement, legal determination, or production deployment is implied.

AI capability, workforce training, continuing education, human review

AI Training Product Design

Problem

Most AI training teaches people how to get a better answer from a model. That is not enough. People also need to know how to read AI output, change the prompt with intent, and edit the result before using it.

Foryn contribution

Foryn developed a training model built around three human skills: READ, PROMPT, and EDIT. The learning experience shows how small changes in direction produce visible changes in the draft, then teaches learners how to inspect and improve the result.

Artifact/system type

Course design, prompt ladder, review framework, practice experience, and continuing-education-ready training structure.

Why it matters

AI capability is not just tool access. It is judgment under real working conditions. Training should help people become better reviewers, not just faster prompt writers.

Status: Product and curriculum design work. Public self-serve training is not currently available.

AI governance, internal controls, review systems, accountable work

Governance and Evidence Workflows

Problem

Many organizations write AI policies before they know how the work will actually happen. The result is a gap between policy language and daily behavior. People use AI, but the organization cannot always see what changed, who reviewed it, or why it was accepted.

Foryn contribution

Foryn developed workflow concepts for use-case classification, required human review, approval blockers, reviewer confirmation, and evidence records. The focus is policy-to-practice translation: turning governance principles into steps people can follow.

Artifact/system type

Use-case classification model, review workflow, approval checkpoint, evidence record, and accountable AI work pattern.

Why it matters

Governance does not become real because a policy exists. It becomes real when the work changes. The human decision layer needs visible steps, clear ownership, and records that survive beyond memory or chat history.

Status: Applied AI governance framework and prototype work. No active public SaaS deployment is implied.

Forecasting, financial technology, signal review, decision support

Predictive Systems Research

Problem

Data-rich environments create pressure to move quickly from signal to action. AI can help with research, screening, summarization, and scenario review, but the process becomes risky when prediction, decision, and execution collapse into one step.

Foryn contribution

Foryn is developing research workflows for AI-assisted forecasting, signal evaluation, and decision support. Current work includes financial-technology research, where the emphasis is on evidence, uncertainty, process discipline, and decision quality.

Artifact/system type

Signal research workflow, forecast memo structure, evidence checklist, uncertainty review, and human decision gate.

Why it matters

Predictive systems should help people reason more clearly under time pressure. They should preserve evidence, expose uncertainty, and separate what the system sees from what a person decides to do.

Status: Internal research and prototype work. This is not investment advice, a trading system claim, or a performance claim.