TEZLA TECHNOLOGIES

Client Portal — AI-usage transparency

← H-Pack Studio

AI Usage Dashboard SAMPLE

Your AI compute at a glance — October 2026 billing period. All figures below are sample data.

AI Active Hours — This Month

0.0 ACTIVE HOURS
19.05 hrs

of an 80-hour monthly AI allotment SAMPLE

Metered at active-compute resolution — idle and monitoring time never counts. How we measure →

Tickets Touched

8
this period · sample

Tasks Completed

83
by AI on your tickets · sample

Resolved

7 / 8
1 in progress · sample

Est. AI Billable

$1,428.75
sample rate · see invoice

Outcome summary ▸

Across 8 tickets this period, AI handled the diagnostic and execution work end-to-end: mail flow restored, 41 mailboxes cleared of a phishing campaign, 14 endpoints hardened to baseline, a print server migrated with zero downtime, and 2,140 stale SharePoint shares remediated. One ticket (VPN tunnel) remains in a 48-hour stability watch.

Our errors are never billed to you. Retries caused by our mistakes, our internal R&D, and idle/monitoring time are metered out before your report is generated. Full methodology →

ACCESS — GATE PENDING

Client sign-in through Cloudflare Access is being connected. Until then, everything shown here is demonstration data — no client information appears on this page.

Tickets SAMPLE

Every ticket our AI touched this period — with the AI active time metered against it. Sample data.

TicketTitleStatus AI active (min)TasksOutcome

AI active time = seconds of real compute on your ticket, accumulated per task cycle. Dispute any row within 15 days — we re-run the ticket log. See How We Measure.

Monthly Usage Report SAMPLE

October 2026 — per-ticket AI usage and billable amounts. These are exactly the rows your invoice is built from.

TicketAI active (min)AI active (hrs) TasksOutcomeBillable
TOTAL
Exports exactly the rows shown above.

Billable AI compute uses a sample rate of $75/hr in this demonstration. Your contract states your actual rate. Rate never changes mid-period.

Invoices SAMPLE

Sample invoices built from the AI usage rows above. Click a header to expand.

INV-2026-1001

▾
Issued2026-09-01
PeriodAugust 2026
Status
Line itemQtyRateAmount
Business managed IT — 12 machines12$175.00$2,100.00
M365 Business Premium — 12 users12$35.00$420.00
AI compute — 8 tickets, 14.2 active hrs (sample)14.2$75.00$1,065.00
TOTAL$3,585.00

AI compute line item is drawn directly from the August usage report. SAMPLE

INV-2026-1002

DUE ▾
Issued2026-10-01
Due2026-10-31
PeriodSeptember 2026
StatusDUE
Line itemQtyRateAmount
Business managed IT — 12 machines12$175.00$2,100.00
M365 Business Premium — 12 users12$35.00$420.00
AI compute — 8 tickets, 19.05 active hrs (sample)19.05$75.00$1,428.75
TOTAL$3,948.75

AI compute line item is drawn directly from this period's usage report. SAMPLE

Service Catalog

Our published rate card — live rates, not sample. What you see here is what you pay.

Managed IT — per machine / month

TierRateIncludes
Essentials$125Monitoring, patching, AV, helpdesk
Business$175Everything in Essentials + backup, email security
Premium Business$225Everything in Business + Command Center, priority response

Professional services — per hour

RoleStandard rateAfter-hours / emergency
Field technician$135$202.50
Senior technician$165$247.50
Engineering / project management$190$285.00

After-hours, emergency, and stabilize-only work is billed at the standard rate + 50% (1.5×).

Microsoft 365 resale — per user / month

LicenseRate
Business Premium$35
Business Standard$25

How We Measure AI Time

Plain language. No black box. If you can read this page, you can audit your bill.

What "AI active time" actually is ▸

When our AI works a ticket, every task cycle is timed in active compute seconds — the seconds the system is genuinely working: reading logs, running diagnostics, executing fixes, writing summaries. The timer runs only while work happens. It stops the moment the system waits, idles, or watches.

In a 5-hour live test, the metering recorded 18,009.1 cumulative active seconds across 398 cycles, metered out 148 errors, and logged every one of the 79 final-check decisions. The same metering that ran that test runs your tickets.

Per-ticket attribution ▸

Active seconds are stamped to the ticket the work was done for. If the AI spends 12 minutes diagnosing your mail flow, those 12 minutes belong to your mail-flow ticket — never to a pool, never to an average. Your usage report lists every ticket with its own line.

What counts as billable ▸

  • Billable: active compute on your ticket — diagnostics, execution, remediation, documentation of the fix.
  • Never billable: our internal R&D; retries caused by our errors (if the AI makes a mistake, the rework is on us); idle or monitoring time; time the system spent waiting on a human or a third party.

How you can verify ▸

  • Per-ticket log lines — every ticket carries timestamped meter entries. Ask and we send them.
  • This report — the numbers here are the numbers on your invoice. Same source, no translation step.
  • Raw logs — available on request for any billing period, in full.

How to dispute a line item ▸

Reply to the invoice within 15 days and name the line. We re-run the ticket log for that line, show you the meter entries, and correct anything that doesn't hold up. No forms, no runaround — a human answers.

The money flow

01

Ticket
Your issue enters the queue.

→
02

AI time metered
Active compute seconds stamped to the ticket.

→
03

Usage report
Per-ticket rows published here.

→
04

Invoice line item
Report rows become the bill.

Ticket → AI time metered → usage report → invoice line item. Four steps, each one inspectable, none of them a black box.

FAQ

Straight answers about how AI time, billing, and your data work.

What is "AI active time"? ▾
The seconds our AI systems spend genuinely working on your ticket — diagnosing, executing, fixing, documenting. Idle time, waiting time, and monitoring time are metered out and never counted. See the How We Measure tab for the full methodology, including our 5-hour live metering test.
Why am I billed for AI time at all? ▾
Because it's real work that used to take billed human hours. The difference: you see every minute of it, per ticket, before it ever reaches an invoice — and AI active time costs far less per hour than human labor. We sell what we've proven out internally first.
What if the AI makes a mistake? ▾
Our errors are never billed to you. If the AI goes down a wrong path, every second of the rework is metered out of your report. You pay for work that moved your ticket forward — nothing else.
Where does the usage data come from? ▾
From the same metering that runs our own operations: per-task-cycle active-compute timers, stamped to your ticket ID. The usage report on the Reports tab and your invoice line items are generated from the same source — there is no translation step where numbers can drift.
Can I get the raw logs? ▾
Yes — for any billing period, in full. Per-ticket timestamped meter entries are available on request, and we'll walk you through them line by line if you want.
How do I dispute a line item? ▾
Reply to the invoice within 15 days and name the line. We re-run the ticket log for that line, show you the meter entries, and correct anything that doesn't hold up. A human answers — no forms, no runaround.
Do you store my data? ▾
Only what the work requires. We keep the ticket record and its metered log lines — the minimum needed to do the work and prove the bill. We don't retain your files, mailboxes, or credentials beyond what's required for the active ticket, and nothing is used to train anything.
Who do I contact with questions? ▾
Louis Henschel, founder — lhenschel@tezlagroup.com. Billing questions, log requests, disputes, or just "explain this line to me" — all go to the same inbox, answered by a human.