Find the right starting point
We identify where AI can actually help, map one workflow with its systems, rules, data and access, check the risk and regulatory requirements, and agree on a pilot with a result you can measure.
See a sample deliverablePut AI to work inFinancial Aid.
For university Financial Aid and enrollment operations. We map the work, improve your systems, and apply AI with clear controls and human review.
Founded by a former Vanderbilt Financial Aid and PeopleSoft professional.
What we deliver
From the first use case to a team that owns it.
Three services, one goal: AI that works inside your real operations, with clear controls and a result you can measure.
Fixed fee, starting at $7,500. About 2–4 weeks.
Implement
We configure the platform you choose—or the one you already have—connect it to your systems and data, redesign the workflow around it, set permissions and controls, and test the exceptions, not just the happy path.
Train and support
Your staff learn on their own work. We leave reusable instructions and documentation, then keep the system current as models, vendors and regulations change—and keep measuring the result.
Most AI pilots stall between the demo and daily work.
The model is the easy part. The value comes from choosing the right work, redesigning it, connecting it securely, and measuring what changed. That is the work we do.
- 01
Find the valuable uses
Look across the work for the tasks where AI saves real time or reduces real risk, and rank them honestly.
- 02
Check risk and regulation
FERPA, GLBA, client confidentiality, institutional policy and vendor terms decide what AI may touch, and how.
- 03
Redesign the workflow
Decide what is automated, what is assisted, and where a person makes the call. Then document it.
- 04
Set governance and controls
Permissions, review points, logging and a clear owner for every decision the system supports.
- 05
Connect your systems and data
Ground AI in the documents, records and applications your team already relies on, with access limited to what's needed.
- 06
Train, measure, keep current
Staff learn on their own work. We track the outcome and update the system as models, vendors and rules change.
Your platform. We make it work.
You may already have one AI agreement or several. We work with what you have approved, and when you are still choosing, the recommendation follows your requirements, data and budget.
- ChatGPT & Codex
- Claude & Claude Code
- Google Gemini
- Microsoft Copilot
- Grok
- Cohere
- Open & local models
Cloud
The fastest start, using the AI services your organization has approved.
Private cloud
Models in your own cloud tenancy, under your security and data agreements.
On-premises & local
Open models running on hardware you control, for data that should never leave the building.
Higher education
Financial Aid and enrollment, workflow by workflow.
We start with one Financial Aid workflow and follow its handoffs to Admissions, the Registrar and Student Accounts. Each map names the systems, rules, exceptions and people who make the decisions.
- 01
ISIR loading and corrections
Map imports, matching, suspense and correction queues. Staff review exceptions and approve record changes.
- 02
Verification
Track requested documents, missing items and conflicting information. A Financial Aid professional completes the review.
- 03
SAP appeals
Organize Satisfactory Academic Progress appeal packets and follow-ups. The institution's designated reviewers decide.
- 04
R2T4
Map withdrawal notifications, Return of Title IV reviews and deadlines across Financial Aid, the Registrar and Student Accounts.
- 05
COA and packaging updates
Document Cost of Attendance changes, award rules and approval points before updating budgets or packages.
- 06
FAFSA and regulatory change
Trace new guidance to the award year, system settings, procedures and staff training it affects.
The Koru Method
Improve the work. Then choose the technology.
Before we automate anything, we learn how your organization actually operates: the rules, the exceptions, the workarounds and the people who know them. AI is applied only where it makes the system better.
The Koru Method04 stages
- 01Understand the real work
Capture the process, documentation, controls, exceptions, workarounds and tacit knowledge—how the organization actually operates.
- 02Improve the system
Remove friction, strengthen the documentation, clarify decisions, and simplify what should not be automated.
- 03Build what earns its place
Apply AI, automation, integration or an agent only where it creates durable value.
- 04Keep the work current
Monitor quality and change, update the operating materials, and measure the outcome over time.
Agents do the groundwork. A person signs off.
We use the same approach on our own business that we bring to yours: supervised agents do the repetitive work, and a person approves anything a client or the public will see. What we run for ourselves, we can set up for you.
Drafts the writing
Agents turn podcast episodes and field notes into blog posts and social drafts, and track regulatory and vendor change.
Keeps conversations moving
Agents sort the inbox, summarize what came in and draft replies. Nothing is sent until a person approves it.
Keeps the lights on
Agents propose site updates as reviewable changes and prepare a daily brief: new conversations, open tasks, spend.
Know which rule supports the answer.
Our Financial Aid reference project maps federal Title IV statutes, regulations and official guidance so a question can be traced to its source, its dates and the workflow it affects. It is the Koru Method in practice: understand how the work operates before automating it.
In development · not yet a public service
See our Higher Ed work
Latest
Notes from the field.
What we are learning about putting AI to work, written up on the blog.
Start with one workflow: mapping Financial Aid before adding AI
Before an assistant touches a single ISIR, map the work: the steps, the systems, the rules, the exceptions and the people who decide.
ReadWhat a global AI alliance tells a university office about AI
PwC and Cohere just split enterprise AI into two jobs: the platform, and everything it takes to make the platform useful. Smaller organizations need the second job too.
ReadWhy Koru runs on agents, and where a person still signs off
We run our own business on supervised AI agents. Here is what they do, what they are not allowed to do, and why that matters for clients.
ReadThree words. No fine print.
Technology changes quickly. Character should not. These principles shape what we recommend, how we build, and how we treat people.
Honesty
If automation is not the right answer, we will say so. Every recommendation includes the limits, risks and evidence behind it.
Quality
A demo is not a production system. We test the work, document the decisions, and design for the exception—not only the happy path.
Friendship
We work beside your people, share what we know, and leave the team more capable. Good technology should feel like support, not displacement.
The goal is not to put AI everywhere. It is to give people time back, reduce the weight of complicated work, and help teams serve others better. If we cannot do that honestly and well, we should not build it.
Your first Koru project
What is one process
your people should not have to fight?
Tell us about one workflow, including the messy parts. We will tell you honestly where AI fits, where it doesn’t, and what a first pilot would look like.
Book a 20-minute callOr email hello@thekoruproject.com