01 — Statement

AI for the agencies that have to deliver.

For state and local government. We build focused AI tooling for agencies and departments — the offices that have to make the call, answer the constituent, sign the filing, file the report. We don't bill for time. We bill for outcomes — measured against the metric in your SOW.

Let's talk → Outcomes-priced · principal-led · federal research backing

$13M+

in federal research backing

across NOAA, the Department of Defense, and other federal agencies — work led by JHG's principals.

15+ yrs

production AI for federal customers

the systems built where reliability, evidence, and audit trails were non-negotiable.

retainer minimums

engagement fees tied to outcomes. Fixed-price discovery up front. You pay when the work works.

02 — What we build

Operational AI, applied to the work your team actually does.

Not a platform. Not a chatbot. A focused tool that fits into your agency's workflow and takes a specific piece of work off your team's plate. Some of these we've built directly. Some we'd build for you, applying the same methodology applied through fifteen years of federal research.

Cities & Counties

Constituent inquiries, answered the same day.

Permit questions, service requests, records requests — drafted and ready for review.

Routine inquiries triaged, drafted, and ready for staff review. The hard ones still go to a person — and arrive on their desk pre-summarized. Same-day response by default.

24h SLA

same-day response by default

State Agencies

Filing review & compliance analysis.

Rule-by-rule analysis. Overnight, not weeks.

The work that used to wait three weeks for the senior reviewer to get to it gets done in a single overnight pass. The reviewer focuses on the judgment calls, not the wording.

3wk → overnight

per filing cycle

All agencies

Public records, fulfilled by the deadline.

Search, draft redactions, bundle. Audit log included.

Records search across systems, draft redactions for sensitive content, bundled response packages with a full audit log. Staff approves; staff doesn't transcribe. Statutory deadlines met without an all-hands fire drill.

21d → 3d

per FOIA / public records request

State & Local Agencies

The grant draft, ready for the program officer.

Pulls from prior funded proposals. First draft in a paragraph.

Paste the RFP and the topic; get a draft pulling from prior funded proposals, your data, and the agency's history with the funder. Quarterly reporting? Same idea, on the cycle the funder requires.

1¶ → first draft

per proposal · per report

Cities, Counties, Agencies

Case documentation, drafted from the visit.

The report's done before the inspector gets back to the office.

Inspector visits the site, dictates notes from the truck. By the time they're back at the desk, the case file is drafted, the photo log is tagged, the next steps are flagged. The inspector edits — not transcribes.

back-at-desk

documentation done · audit-ready

Cities & Counties

Permit applications, pre-reviewed overnight.

Routine ones approved next morning. Complex ones flagged with the issue.

Applications come in. The system checks completeness, validates against ordinance, drafts the response. Staff approves the easy ones and focuses on the gray areas — with the relevant code section already cited.

next-day

routine permits · pre-cited reviews

If one of these sounds like a problem your office is solving the hard way, that's a conversation we'd like to have.

03 — Pricing & process

We charge for outcomes. Not for time.

Most consulting and platform vendors bill you whether their work moves your numbers or not. We don't. A short fixed-price discovery to figure out what's worth building. Then a defined-scope build with an outcome-tied engagement fee — measured against the metric we agree to in advance. If the lift isn't there, we don't keep billing.

Pricing

Fixed-price discovery up front. Then outcome-tied engagement fees for the build. Real skin in the game — the metric is named in your SOW.

↳ 01

Discover.

2 weeks · fixed price · plan + price + metric

We sit with your team. Map the workflow, the data, the decision the tool has to support, the policy and privacy posture. You walk away with a plan, a scope, a price, and the metric we'd tie engagement fees to — even if you don't hire us.

↳ 02

Build to the metric.

every 2 weeks · production-grade increments

Each cycle ends with a working tool your team can actually use. We're shipping against the metric, not delivering slides. The smallest thing that's actually useful, before we add the next thing.

↳ 03

Outcome pricing kicks in.

paid against the metric · measured monthly

Once the tool is in production, our engagement fees track the metric we agreed to. The better it works, the more we make. The less it works, the faster you find out — and we adjust scope, no penalty.

↳ 04

Procurement-ready.

defined SOW · per-unit fee · audit trail

Take the SOW to your procurement office. The metric, the per-unit engagement fee, the deliverable schedule, and the data-handling posture are all named. Your contracts office gets a contract they recognize — performance-based, with a defined ceiling.

04 — Track record

Built for customers who needed it to work.

Fifteen years of production AI for federal customers — NOAA, the Department of Defense, and the agencies that hold work to evidentiary standards. Over $13 million in federal research funding led by the principals across that span. The same engineering rigor — methodology, data discipline, decision-support delivery, audit-ready artifacts — applied to the state and local agencies that have to deliver, not just propose.

Operational AI is what happens after the model works in the lab. It's the integration, the workflow fit, the audit trail, the part you can't write a paper about.

— a JHG principle

Track record

2000s

Research foundations: machine learning methodology in security, decision support, and data systems for federal customers.
2010s

Production AI at federal scale.

Defense R&D · homeland-security analytics · mobile and embedded systems for first responders · the systems agencies bet on.
2010s — early 2020s

Federal research portfolio: $5M+ from NOAA, $13M+ across federal agencies — led by the principals.

Methodology and delivery, in the domains where evidence is graded and the work has to hold up under review.
2025

Jones Hill Group founded.

Principal-led, outcome-priced engagements bringing the engineering rigor of federal research to the agencies that have to deliver.
now

State and local government — agencies and departments across the country.

05 — Posture

Compliance, audit trail, and the way we work with your existing systems.

Public-sector buyers have learned to read between the slide-deck lines. The questions a procurement officer or agency CIO actually asks have answers below — explicit, scoped, and named in the SOW.

Compliance

State IT security standards, IRS Pub 1075 where applicable, public-records preservation.

Engagements are designed around the data privacy and security regimes your agency actually operates under — state IT standards, CJIS where relevant, IRS 1075 for tax data, public-records and retention statutes. The discovery phase names the applicable standards explicitly; the build is scoped to meet them.

Data residency

Your data stays in your environment.

We deploy onto your infrastructure or one you control — government cloud, on-premise, or your existing approved platform. No JHG-hosted cloud holding your records, no outbound data feeds to third parties. The data doesn't move unless you say so.

Audit trail

Every decision logged. Every output explainable.

The tool keeps a complete log of inputs, decisions, and outputs — what the system saw, what it returned, who reviewed it, and when. Audit-ready by default. Public-records-ready by design.

06 — Integrations

Built to work with the tools you already use.

We don't replace your existing systems — your case-management, your records management, your ERP, your permitting platform. The AI tooling fits into what you already run, from Tyler Technologies, Granicus, OpenGov, and ServiceNow on the institutional side to the everyday productivity tools your team uses every day:

Don't see what your office uses? Day one of discovery is asking what's in your tools and what we'd integrate with.