01 · Digital product builders · est. 2022

Build & Ship quickly

JMS Analytics is a small studio that leverages AI to ship comprehensive and valuable digital solutions without cutting corners. The rigor of a long engagement, the pace of a sprint.

PracticesBusiness intelligence·Advanced analytics·Custom software
02 · What we build

A single toolkit. Many shapes.

Business intelligence

What the business actually looks like

Dashboards, reporting layers, and the semantic models behind them. Descriptive products that turn scattered data into a clear view of what is and what has been. The kind of clarity that makes the next decision obvious.

Tools
Power BIDAXTMDLMSQLHexStreamlitPlotlyMS Fabric

Advanced analytics

What's likely to happen next

Forecasting, segmentation, classification, ad-hoc modeling. Forward-looking products that move beyond reporting into prediction, and where useful, prescription. Delivered as production-ready apps your team can run on a schedule, not notebooks that bit-rot in a drawer.

Tools
PythonScikit-learnProphetNixtlaPyTorchPySparkHex

Custom software

The software that runs the work

Full-stack web applications, internal tools, customer-facing apps, AI-powered interfaces. Often where business intelligence and advanced analytics live in production: wrapped in software that earns its place in daily operations.

Tools
TypeScriptNext.jsReactTailwindshadcn/uiPostgreSQLVercelSupabase
03 · Recent work

Real shipped solutions.

See all work
Business intelligenceCase 01
Clientlocal government

Project Portfolio Status Reporting

Replaced client's manually compiled Excel portfolio status reports with a governed Microsoft Fabric Data App that gives eight departments at-a-glance triage of their Portfolio, refreshed straight from their application database.

OutcomeReplaced hand-built Excel status reporting with a governed, self-refreshing status dashboard, saving hours of manual effort per year and serving 60 people across the organization
DurationJun 2026 – Aug 2026
Featured
Advanced analyticsCase 02
Clientlocal government

Customer Support Service Topic Analytics & FAQ

A pipeline that ingests and processes IT Customer Support Service data, discovers and assigns topics to each ticket, ranks them by volume and resolution time, and produces a FAQ for staff to review on demand.

OutcomeTopic modeling and discovery automation saves hundreds of hours per turn and equips managers with topical low-hanging fruit for efficiency gains
DurationMay 2026 – Jun 2026
Custom softwareCase 03
Clientlocal government

Income Tax Q&A Agent

Designed and built an AI assistant that answers client staff and tax administrators' questions about the income tax, replacing hours of repetitive phone and email support that spike around quarterly filing deadlines.

OutcomeMoved hours of repetitive phone and email questions on to an AI-based Q&A agent that answers quickly, accurately and consistently, saving hours of staff time
DurationNov 2025 – Aug 2026
Business intelligenceCase 04
Clientlocal government

Forecast Accuracy Tracker

A monitoring and diagnostic report that tells client how much to trust its demand forecasts - measuring accuracy and bias against arriving actuals across vintages, horizons, and methodology eras, which replaces roughly eight hours of manual forecast tracking each month.

OutcomeReplaced ~8 hours/month of manual forecast tracking with an automated BI accuracy monitor
DurationApr 2026 – Jul 2026
04 · How we work

We start with a conversation - not a contract.

Intake and a written proposal are on us. Once we agree on scope, we build with AI in the loop at three or four times the pace of a traditional consultancy without skipping the things that make a system worth owning. Most engagements run six to twelve weeks.

00On us
Intake and proposal
A call to understand the problem and a written proposal back. No theatrical kickoff workshop, no charge to figure out whether we're a fit.
01
Requirements and design
We read your codebase, your docs, your data, and we talk to the people who actually use the system. You get a written requirements doc and the shape of the thing: data models, interface mockups, API contracts, whatever matters for your project. Cheap to change now, expensive to change later. You sign off before we build.
02
Build and test
We build with AI tooling in the loop for code generation, automated review, and AI-assisted data exploration, and we test as we go. Tests, observability, a runbook. We don't ship the happy path and leave. Edge cases, failure modes, and the things only your team knows about get covered.
03
Deploy and handoff
Production deployment, architecture docs, video walkthroughs, and pairing sessions with your engineers. The goal isn't to keep you on a retainer; it's to leave a system your team can confidently own.
05 · Contact

Have a project in mind? Let's talk.