What is this course about?
Budget work is high-volume document analysis attached to numbers that must be exactly right. This course covers where AI helps across the budget cycle, grant solicitation and application screening, subrecipient monitoring, improper payment detection and its false-positive cost, and the numeric verification gate that prevents a generated figure being published.
Who is this course for?
- Budget analysts and examiners
- Grants management officers
- Program integrity and improper payment staff
- Internal audit and inspector general staff
What do I need before starting?
- Familiarity with the budget or grants cycle in your agency
- No technical background required
What will I be able to do afterwards?
- Identify where AI helps in the budget cycle and where it must not
- Analyze appropriations and spending across fiscal years
- Screen grant applications with documented, defensible criteria
- Run subrecipient monitoring at portfolio scale
- Enforce a numeric verification gate before any figure is published
What does each module cover?
Where does AI help in the budget cycle?
40 minMapping the cycle to find the document-heavy work AI accelerates and the decisions it must not touch.
Objectives
- Map the budget cycle by task type
- Identify document-heavy versus decision tasks
- Set the boundary before deployment
Topics
Activity. Map your budget cycle and classify every task by AI suitability.
How do you analyze spending across fiscal years?
50 minMulti-year appropriations analysis where account structures and definitions change.
Objectives
- Analyze spending across changing account structures
- Handle definitional changes between years
- Detect and explain anomalies
Topics
Activity. Analyze three years of spending across an account structure change.
How do you screen grant applications defensibly?
50 minApplication screening with criteria that can be documented and defended to an unsuccessful applicant.
Objectives
- Apply documented screening criteria consistently
- Keep the funding decision with human reviewers
- Document reasoning for every screening outcome
Topics
Activity. Screen a set of applications and document the reasoning for each outcome.
How do you analyze a solicitation for applicants?
45 minHelping applicants and program staff extract requirements from a complex funding opportunity.
Objectives
- Extract requirements from a solicitation completely
- Verify extraction against the source
- Generate applicant-facing checklists
Topics
Activity. Extract and verify every requirement from a real solicitation.
How do you monitor subrecipients at scale?
45 minPortfolio-scale monitoring and single audit preparation.
Objectives
- Design portfolio-scale monitoring
- Prioritize by risk rather than by size
- Prepare single audit documentation
Topics
Activity. Build a risk-based monitoring plan for a real subrecipient portfolio.
What does a false positive cost in improper payments?
45 minDetection where flagging a legitimate payment has a real cost to a real recipient.
Objectives
- Model the cost of false positives to recipients
- Tune detection against that cost
- Design a review path before any action
Topics
Activity. Model false positive cost for a detection scenario and tune accordingly.
How do you gate every published number?
45 minThe verification gate that makes it structurally impossible to publish a generated figure.
Objectives
- Separate generated narrative from computed figures
- Implement a verification gate before publication
- Trace every published figure to its calculation
Topics
Activity. Implement the gate and attempt to publish an unverified figure.
Building the grant analysis agent
50 minThe lab module: an analysis agent with a non-bypassable numeric verification gate.
Objectives
- Build the analysis agent
- Integrate the verification gate
- Produce an audit trail for every analysis
Topics
Activity. Build the agent and verify no figure can reach output without passing the gate.
What is the capstone project?
Grant analysis agent with a numeric verification gate
Build a budget or grants analysis workflow with multi-year spending analysis, defensible application screening, verified solicitation extraction, risk-based subrecipient monitoring, false-positive-aware detection, and a non-bypassable numeric verification gate.
Deliverable: A working analysis agent with a demonstrated verification gate and full audit trail.
How are learners assessed?
- Verification gate tested — publishing an unverified figure must be impossible
- Solicitation extraction verified complete against the source document
- Screening reasoning reviewed for defensibility to an unsuccessful applicant
What ships with the course?
Facilitator guide
Session-by-session running order, discussion prompts, and the questions that reliably derail a room.
Learner workbook
Exercises, checklists, and the templates each module's activity produces.
Hands-on lab environment
A sandboxed ibl.ai deployment so exercises run against real agents, not screenshots.
Assessment bank
Scenario questions and rubric criteria mapped to each stated learning outcome.
Source bibliography
Every primary regulation and standard cited on this page, linked and dated.
Which AI agents does this course use?
The hands-on modules run against agents already deployable on the ibl.ai platform for government.
Where does the course material come from?
Every module is grounded in primary sources — the regulation, standard, or research itself, not a summary of it. Each was resolved at authoring time.
- Government Accountability Office
GAO
Audit standards and improper payment methodology.
- Office of Management and Budget
OMB
Uniform Guidance and federal financial management policy.
- Federal Acquisition Regulation
Acquisition.gov
Acquisition rules intersecting with grants and payments.
- AI Risk Management Framework
NIST
Risk framing for the detection and screening modules.
Delivery notes
Binding guidance for anyone preparing and delivering this course.
- Module 7's gate must be structural. Budget offices under deadline pressure will bypass a soft control, and a published generated figure is a public credibility failure that is hard to recover from.
- Module 6's false positive framing is the ethical core. A flagged legitimate payment can cause real hardship to a recipient, and detection tuned only for recall causes harm.
- Module 3 must keep the funding decision human. Screening for completeness and eligibility is administrative; ranking for award is not, and the line should be explicit.
- Use real solicitations from at least two agencies. Structures differ enough that a single-agency example does not transfer.
- Coordinate with GOV-4 — solicitation analysis appears in both, and the extraction methodology should be shared rather than taught twice.
Why run AI training on a platform you own?
You own the course, not a licence to it
Course content, learner data, and the platform run inside your perimeter — you own all the code and the data.
Model-agnostic delivery
Run the course's AI components on any LLM — Claude, GPT, Llama, Gemini, Command — and switch anytime.
No per-seat training licences
Usage-based or self-hosted, so cost tracks actual use rather than headcount.
Deploy anywhere
Cloud, private VPC, on-premise, or fully air-gapped — including for cohorts that cannot use public AI tools.
Frequently asked questions
What does the Budget and Grant Analysis with AI Agents course cover?
Budget work is high-volume document analysis attached to numbers that must be exactly right. This course covers where AI helps across the budget cycle, grant solicitation and application screening, subrecipient monitoring, improper payment detection and its false-positive cost, and the numeric verification gate that prevents a generated figure being published. It runs 5 hours across 8 modules across 8 modules, at intermediate level, and closes with a capstone: Grant analysis agent with a numeric verification gate.
Who should take Budget and Grant Analysis with AI Agents?
It is written for Budget analysts and examiners, Grants management officers, Program integrity and improper payment staff, Internal audit and inspector general staff. Prerequisites: Familiarity with the budget or grants cycle in your agency; No technical background required.
Can we run this course on our own infrastructure?
Yes. ibl.ai is model-agnostic and deploy-anywhere — cloud, private VPC, on-premise, or fully air-gapped — and you own all the code and the data. Cohort data, submissions, and any material learners upload stay inside your perimeter, which matters for government teams that cannot send work to a public AI tool.
How do we get access to Budget and Grant Analysis with AI Agents?
Request access and we will set it up for your cohort — hosted by ibl.ai, or running against your own deployment. Tell us the group size and timing you need, and whether it should run inside your own perimeter.
How much does AI training for government cost on ibl.ai?
There is no per-seat pricing — you pay for usage or self-host and pay only for the infrastructure, so a 5,000-person rollout does not cost 5,000 licences. 1.6M+ users across 400+ organizations run the platform this way, including NVIDIA, MIT, and Syracuse University.