Nonprofit Grant Management Software, Now With AI

An auditor picks your largest federal program and asks what one case manager actually worked on in March. Then she asks you to account for the rest of that person’s time as well.

TimeRewards Grant Management is built so both answers come out of one record, and it is available now. You upload the agreement, the approved budget, the terms, the subawards and the amendments. AI reads them and proposes the award facts for somebody on your team to accept, edit or reject. From then on effort, match, indirect recovery and every reporting deadline sit on the award itself.

The rest of this post is what that looks like across a grant year. What the AI reads, what it leaves to your people, and what changes for a nonprofit running one workforce across several funders. Every regulation quoted is read from 2 CFR Part 200.

The year usually starts well. The award letter arrives and the finance director opens the terms. By the end of the week she has a list. Approved budget categories, the match you committed to, the indirect rate, the reporting schedule, and the partners you now have to monitor. Then the program starts. A case manager splits her week between three funders, and a van is shared by two programs. A budget line turns out to be wrong by February.

What is TimeRewards nonprofit grant management software?

Nonprofit grant management software, in our hands, means an award record that sits under your timesheets. Most time tracking starts with a project list somebody typed in. This starts with the award. The codes your staff pick are connected to the funder, the budget category and the restrictions that award carries.

The real difference is where the rules live. In most organizations they live in a terms document, a spreadsheet of budget lines and the finance director’s memory. All three are consulted at month end. Held against the award instead, they reach the moment an hour is recorded, which is the only moment they can still change anything.

Dates matter as much. Budget periods, rates, allocation bases and deadlines all carry them. An amendment changes what happens next without erasing what was true before it.

So the unit of work changes. Instead of running programs that happen to be grant-funded, you run awards, and the programs hang off them.

What does the AI read, and what does it propose?

It reads the documents you upload and proposes the facts it finds in them. Each proposal is a field with a value and a source, waiting for a person to accept it, edit it or reject it.

The sequence is the same for every award, and it is short.

1. Upload

   The agreement, the approved budget, the terms, the subawards and every amendment. They stay attached to the award record, so the document behind a figure is one click away when an auditor asks about it.

2. AI proposes

   TimeRewards AI reads the documents and puts forward the award facts it finds. Nothing is adopted at this point. A proposal is a suggestion with a document behind it.

3. A person accepts, edits or rejects

   Someone on your team works through the proposals, and the reviewer is recorded against each decision. That is what turns an extracted figure into a fact your organization stands behind.

4. Connect the award to the work

   Programs, activities and match work get their codes, and the funder’s restrictions attach to them. From then on the limits are visible where hours are entered, rather than in a terms document nobody reopens.

   Set up once per award. Staff carry on picking the same programs and activities they already know.

5. Work, and let the record build

   Effort is recorded as it happens and routed for certification, and match accumulates in its register. The indirect base recalculates as costs post, and amendments arrive for review rather than changing anything silently.

Most of the effort is in the first hour of an award’s life. After that the record maintains itself, because the things that change arrive as amendments and each one comes back to a person.

What does it change for a grant-funded nonprofit?

Nine things, and they arrive in the order a grant year runs. Each one used to be somebody’s spreadsheet.

Effort that accounts for all of someone’s time

Start with the standard everything else is measured against, because it is stricter than most people expect.

The same paragraph requires that the records cover the total activity the person is paid for, capped at 100 percent. Budget estimates alone do not qualify as support. So the requirement is not that you document the 45 percent charged to one grant. It is that you can account for all of that person’s time.

The federal share is the smaller claim. The harder one is the full account, including the fundraising hours no federal award may pay for. The names and figures in these examples are illustrative.

In TimeRewards the effort record is built as the work happens, against the funder and the activity it supports. It is then routed for certification by somebody who knows what the person did. Where a budget estimate was used for interim charging, the variance against actual effort is visible and the adjustment sits on the record.

What makes that possible is the award sitting underneath. The activities staff pick are tied to budget categories and funders the AI proposed and your team accepted. So time and effort reporting stops being a quarterly reconstruction and becomes a review of what was already recorded.

Many funders, one team, one set of rules at the point of entry

Each funder writes its own rules, and the same person works under all of them in a single week.

When those rules live in four places, the reconciliation happens at month end in a spreadsheet, by one person, from memory.

Grant Management holds each funder’s restrictions on the award record and applies them where time is recorded. The limit is visible when the charge is made rather than three weeks later. Unlimited Projects and Tasks on every plan gives each award, program and match activity its own code. Automated Email Reminders keep entries arriving while the week is still fresh.

Two budgets for one award, side by side

They answer different questions. The approved budget is a compliance document. It sets the categories you may spend in and the figures your financial report will be compared against.

Your operating plan is a management document. It might hold back contingency, phase hiring into January, or organize costs by activity instead of by category. Force one to do the other’s job and program managers end up working from numbers the funder never approved. Or finance ends up managing a program through categories that mean nothing on the ground.

The approved budget is one of the documents the AI reads. It proposes each category and its amount, and somebody checks them against the award. Your operating plan sits beside them as your own numbers.

The $16,000 gap in personnel is not an error. It is a decision somebody made in October about hiring. Both budgets sit on the award record with the distance between them showing. The conversation about a budget revision then starts from evidence rather than surprise.

Indirect recovery, on a base that recalculates

Nonprofits usually lose indirect money through the base rather than the rate. An organization without a negotiated rate can use the de minimis rate, and the rate is the simple part.

The base is the harder part. Under 2 CFR 200.1, modified total direct costs leave several things out. Equipment, capital expenditures, rental costs, tuition remission, scholarships and fellowships, participant support costs and patient care. They count only the first $50,000 of each subaward.

Every new subaward changes that base, and so does every piece of equipment you buy. Work it out once at proposal time and never again, and it drifts. Drift in the generous direction becomes questioned costs. Drift in the cautious direction, which is more common, quietly leaves money you were entitled to on the table.

The indirect rate and its base come off the terms in the first reading pass. The AI proposes both, your team confirms them, and the exclusions and effective dates sit on the award record beside them.

From there the base recalculates as costs post. Equipment bought in March and a subaward signed in June move it, without anybody reworking a spreadsheet. The figure on your financial report can be reproduced rather than reconstructed.

Match tracked from month one, not at closeout

Cost sharing counts when it is verifiable in your records and allowable under the same cost principles as the federal share. It also cannot be counted toward another federal award. For staff time the rule is explicit.

Match is also where nonprofits track least, because it arrives as volunteer hours, donated space and other funders’ money. None of that flows naturally through payroll or accounts payable.

Take an award with a $412,000 federal share and a 25 percent match, which is $103,000. Eight months in, $78,000 is documented: $48,000 in cash from unrestricted funds, $21,000 in volunteer hours and $9,000 in donated space. The $25,000 gap needs $6,250 a month for the last four months. Found in July, that is a planning problem. Found at closeout, it can reduce the federal share you keep.

The commitment itself is proposed off the agreement, along with how in-kind is to be treated, and confirmed by whoever reviews the award. The match register then tracks that commitment against documentation, month by month.

Cash, volunteer hours and in-kind contributions sit against the award they count toward. The gap shows while there is still time to close it, rather than at closeout.

Shared costs, with a basis and a review date

Rent, utilities, the database license and the finance director’s salary belong to no single award. They need a method you can explain, applied consistently, and revisited when the thing it is based on changes.

Where proportions cannot be measured because the work is too intertwined, 2 CFR 200.405(d) allows allocation on any reasonable documented basis. The documentation is what makes it reasonable. A square-footage split holds up if you can produce the floor plan it came from.

The last row is the one to fix. Nothing about it looks wrong in the general ledger, and it is running on a headcount almost two years old. TimeRewards holds each allocation basis with the date it was last reviewed, and surfaces a stale one before somebody requests the workpapers.

Subawards monitored, not just issued

Whether a partner is a subrecipient or a contractor depends on what the relationship does, not what the agreement calls it. Under 2 CFR 200.331, the substance of the relationship matters more than the form of the agreement.

A subrecipient carries out part of your federal program and has its performance measured against the program’s objectives. A contractor provides goods or services in the ordinary course of its business. The difference decides what you owe.

For a subrecipient, you verify it is not excluded, issue a subaward carrying the required information, assess its risk and monitor it accordingly. Where your indirect base is modified total direct costs, only the first $50,000 of each subaward counts toward it.

Subawards go in with the rest of the documents, and the AI proposes each partner, its amount and the monitoring the award expects. From there the subaward monitoring register keeps each partner’s risk tier, the monitoring you committed to and the date it was last done.

An overdue desk review shows up while the partner is still invoicing you, which is the only time it is cheap to fix.

Deadlines that move when the period of performance moves

A printed date is easy. A date calculated from something that has not happened yet is not, and those are the ones that cause trouble.

Closeout is the clearest example. Under 2 CFR 200.344, a recipient submits all final reports no later than 120 calendar days after the period of performance ends. A subrecipient reports to its pass-through entity within 90. Records are then kept for three years from the date you submit the final financial report. Both clocks start from an event rather than a date.

An extension moves every one of those dates. Your award terms may authorize a one-time extension of up to 12 months. 2 CFR 200.308 then requires written notice at least 10 calendar days before the period of performance ends. The extension cannot be used only to spend remaining funds.

Budget transfers work the same way. An agency can restrict transfers among direct cost categories when the federal share is above the simplified acquisition threshold. Where it does, the test is cumulative: once transfers pass 10 percent of the total budget, including cost share, you need approval first.

No single move crossed the line. Together they did. The running total is tracked against the approved budget, and the proposed transfer is flagged before the money moves. At that point a request is a short email rather than a finding.

Reporting obligations sit on the award with an owner each, and deadlines are held as a rule plus an event. So an extension reschedules every reminder attached to it, instead of leaving four calendar entries quietly wrong.

Single Audit readiness, as a by-product

An organization that expends $1,000,000 or more in federal awards in its fiscal year needs a single audit under 2 CFR 200.501. The test is on money expended, counted across every federal source, including funds received through state and local pass-through entities.

The audit starts from your schedule of expenditures of federal awards, and then it samples. A few employees, a few transactions, a few months. Show the effort record behind one employee’s March charges. Explain how the occupancy allocation was worked out. Demonstrate that the match on the final financial report was documented.

Each of those takes minutes when the record was built while the work happened. It takes days when it has to be assembled from email and a shared drive.

Grant Management keeps the elements that schedule needs on the award itself: the assistance listing, the awarding or pass-through agency, and expenditure by program. Pulling the schedule then starts from a record rather than a reconstruction, and the evidence behind each line is attached to it.

What the AI leaves to you

Reading an award quickly is the easy half. Deciding what it means for your programs is the half that carries risk, and that half stays with your people.

Each fact the AI proposes waits for somebody to accept it, edit it or reject it, and the record shows who decided. That gap between reading and adopting is where your controls live. A figure extracted wrongly and adopted silently would sit in the record looking exactly like a correct one. A figure a named person accepted is a decision somebody can explain, two years later, to an auditor who was not in the room.

The same applies to compliance itself. No software makes an organization compliant. Compliance lives in your written policies, your separation of duties, your board’s oversight and how consistently people follow them.

A system does something narrower. It makes the correct entry the easy one, shows exceptions while they are cheap to fix, and keeps the evidence attached to the decision it supports.

What setting it up actually takes

Less than people expect, because the work is per award rather than per person.

1. Your first award, about an hour

   Upload the agreement, the budget and the terms, then accept or correct the proposed facts. Give the programs and match activities their codes, and add the funder restrictions and the reporting obligations with owners.

2. The rest of the portfolio, faster

   Every award after the first goes quicker, because the programs and activities already exist. Older awards can go in as they are touched rather than all at once. Start with the awards running this year.

3. Nothing changes for the people entering time

   Staff carry on picking the same programs and activities. The award structure sits underneath, which is the point: a control nobody has to remember is a control that holds.

   Approvals stay where they are. Custom Approval Workflow is on every plan, and Multi-Level Timesheet Approval Workflow on Professional adds the second reviewer for certification.

How to evaluate nonprofit grant management software

Whether you look at ours or somebody else’s, seven questions separate nonprofit grant management software from a document library with a search box. Take them to any demo.

Our answers to all seven are on this page, which is the honest reason to publish a list any nonprofit grant management software should survive. On the last one: nothing the AI reads reaches the record until a named person accepts it, and the record keeps who that was.

What should you check this month?

Pick your largest federal award and one person who charges to it. That is enough to test the whole chain from end to end.

Does their record account for all of their paid time, or only the grant-funded share? If a budget percentage is used for interim charging, when was it last compared with what they actually did?

Then take the award itself. Work out the modified total direct cost base from this year’s actual costs and compare it with the base behind your indirect claim. Write down required match against documented match. Add up every transfer among budget categories so far.

Last, find the period of performance end date and count back ten calendar days. If you might need an extension, that date is the one to put on a calendar now.

Every one of those questions becomes a lookup once the award is in TimeRewards, which is the whole argument for putting it there. Running them by hand first is still an afternoon well spent, because it tells you which answers you do not currently have.

For the software side there is our grant compliance time tracking page, and the nonprofits overview covers how grant-funded organizations usually set things up. If you allocate staff by percentage, auto-filling timesheets by percentage covers the mechanics, and the FTE calculator helps check a split against real headcount. The employee cost calculator shows what an hour of a split salary really costs, benefits included. Organizations that also hold federal contracts get the same treatment for line items and labor categories in government contract management software. Our roundup of 2026 nonprofit compliance changes tracks what else is moving.

Questions people actually ask