NEONALSA ETHICA
OBSCURA

AI integration & agent orchestration

AI that works
with your business.

Connect the information you already have to the work you need done.

I design and build AI workflows around emails, CSVs, spreadsheets, documents, and existing business tools. The result: information that is easier to use, decisions with context, and less manual assembly.

Explore the examples
EMAIL CONTENTCSV & EXCELPDF & WORDBUSINESS RECORDSCONNECTED TOOLS

Explore the possibilities

See the work come together.

Three walkthroughs inspired by systems I’ve built. Choose an example, inspect its sources, and try the review step.

Interactive illustrations · Fictional business data · Prepared sample outputs, with no live accounts or AI calls.

01 / AI-ASSISTED INTAKE

A quote changes. Your records keep up.

Bring a supplier’s message together with an existing purchasing record. Review the proposed difference before updating the project.

SELECTED SOURCES02 files
EML
Re: revised regulator quoteFictional supplier message

For the 12 regulators in your request, the revised price is $72.00 each, USD. Please treat this as a quote.

We will confirm the delivery date separately.

Source 01 · paragraph 1–2
CSV
purchasing.csv

Regulators · Qty 12 · $74.00 each · Quote

PROPOSED PROJECT UPDATEReady for review

Regulator purchase

Matched to the existing record

Unit price
$74.00$72.00
Quote total · 12 units
$888.00$864.00
Status
Quote Unchanged
Delivery date
Not confirmed

Price supported by source 01, paragraph 1. Delivery remains open until confirmed.

The existing record stays unchanged until you approve the update.

The business value Keep purchasing information consistent while preserving the difference between a quote and an order.

Built into NEO Project Desk ↗
02 / DATA INTEGRATION

The files become a view you can use.

Bring operational exports and workbook measures into one workspace, with clear source status and exceptions that need attention.

SOURCE CHECKS02 feeds
CSV
work_orders.csv

6 sample rows · 5 ready to include

1 to review
XLSX
weekly_metrics.xlsx

Selected sheet and reporting period

Read
Work-order excerpt · fictional records
RecordStatePriority
WO-201ReadyRoutine
WO-202WaitingHigh
WO-204ReadyMissing
WO-204 needs a priority.

This row is kept out of the accepted set until corrected. The original source remains available.

OPERATIONS AT A GLANCESample data
READY WORK3Accepted records
WAITING2Accepted records
Ready to proceed3
Waiting on a dependency2

1 source row needs review
Flagged records stay visible.

SOURCE CONTEXT

Work orders: CSV export
Weekly measures: workbook
View: selected reporting period

The business value Spend less time assembling exports and make decisions with a clearer view of data quality. These checks use defined rules.

Explore the dashboard design ↗
03 / AI-ASSISTED REPORTING

A report with the evidence attached.

Give AI a selected snapshot of project records and source material. Review the draft, follow its references, and accept the sections you want.

SELECTED REPORT CONTEXTSnapshot
01
Project purchasing record

12 regulators · $72.00 per unit
Quote total: $864.00 · Status: Quote

02
Supplier message

“We will confirm the delivery date separately.”

03
Project task

Confirm delivery before scheduling installation.

The draft uses this selected context. Your working project records remain separate.

MANAGEMENT BRIEFINGDraft section

A clear update.
A visible next step.

The regulator quote totals $864.00 for 12 units. The purchase remains at quote stage. [01]

Delivery is awaiting supplier confirmation. Confirm the date before scheduling installation. [02, 03]

Open item

Delivery date not yet confirmed.

Review the wording and references before accepting a section.

The business value Turn project detail into a readable update while keeping the sources, open questions, and review decisions close at hand.

Explore the project workspace ↗

The expertise behind the workflow

Connected tools.
Coordinated AI.
Practical judgment.

I’m Charles “Colby” Jackson. My work brings together AI integration, custom software, and a background in controls, automation, maintenance leadership, and project coordination.

Meet the person building it ↗
01

Connect the information.

Extract supported files, structure records, and connect approved business tools. My projects include Google Sheets and Drive integration, email workspaces, and SQL-to-CSV reporting pipelines.

02

Orchestrate the work.

Give AI models defined tasks, selected sources, and the tools they need. Validate returned proposals, manage job states, and make provider usage and costs visible.

03

Make the result usable.

Build the interface, review steps, exports, documentation, and handover around the people doing the work. Test representative cases and handle the exceptions.

Confidence through clarity

Useful AI.
You stay in control.

You choose the scope. Agree what the workflow can read, which tools it can use, and what requires approval.

You can follow the result. Keep original sources, references, and reviewed versions available.

You can start small. Begin with one useful workflow, see how it performs, and grow from there.

Questions before you begin

Understand the AI.
Choose with confidence.

Do I need to replace my existing software?

Often, the starting point is a connection between tools you already use. We review available exports, APIs, account permissions, and the way your team works before choosing an approach. Any required software changes are part of the proposal.

Who can access my business information?

Before connecting anything, we agree which information the workflow needs, which accounts and providers will handle it, and who has access. Hosting, retention, and provider terms depend on the chosen setup. The website examples use fictional data; please keep confidential records out of your initial inquiry.

Can I approve changes before they happen?

Yes. Review can be built into the workflow so AI prepares a proposed update or draft and a person decides what to accept. Project Desk demonstrates selected changes and report sections being reviewed before application.

What happens when AI gets something wrong?

The workflow can check required fields, allowed changes, and source references, and flag missing information for review. Those checks support human judgment. We agree representative acceptance examples, exception handling, and how to correct or recover from a problem.

What ongoing costs should I expect?

Professional fees are listed on the Fees & support page. Hosting, model/API usage, and software subscriptions are separate unless your quote says otherwise. We identify the services, account ownership, and expected usage before you commit. Usage visibility and limits can be included in the agreed workflow.

Can we start with something small?

Yes. One report, one document-intake task, or one connection can be a useful first project. A clear task can receive a direct quote. If the problem needs investigation, we can agree a paid diagnostic first.

Will my team know how to use it?

The agreed build includes operating instructions and a walkthrough suited to the tool. Additional staff training, documentation, and ongoing support can be scoped as needed.

Start with your everyday work

What are you still copying
from one system to another?

Tell me about the emails, spreadsheets, or reports taking up your team’s time.

Discuss your AI workflow