Proposal · 28 May 2026

Your 25 years of knowledge,instantly searchable.

An AI knowledge and diagnostics platform for InMed Healthcare. Built on Amplor, deployed your way, learning from every answer your engineers and applications team give.

Prepared forJeff Gibson, InMed Healthcare
Prepared byJoran van der Smagt, JCR Computers
PlatformAmplor, by JCR
Summary

The fastest value, the lowest risk, first.

InMed has built deep knowledge across medical imaging and diagnostics. It lives in service manuals, bulletins, IFUs, your engineers' heads, and thousands of past emails and service records. Today it is hard to search and impossible to learn from at scale.

Amplor changes that. We can deliver every stage on your wish list. Our recommendation is to start where value lands fastest and risk is lowest: your two internal knowledge bases. Both run on the same engine, both stay internal while we tune them, and both give your teams a tool they trust before anything reaches a customer.

YOU CHOOSE, INDEPENDENTLY

Two decisions, your call

  • How it is hosted: cloud (fastest, lowest cost), on-premise on your own VMs (your preference), or a fully self-developed platform you own outright.
  • How the AI handles your data: standard OpenAI, OpenAI Zero Data Retention (recommended for medical data), or fully local AI with no external dependency.
Understanding InMed

Your edge is the support, not just the equipment.

100% Australian-owned, 25 years across X-Ray, MRI, DEXA bone densitometry, nuclear medicine and cardiology. Your differentiator is after-sales: training, compliance testing, custom lab design, and repairs.

That support model is exactly where an AI knowledge base earns its keep. Every fault your engineers diagnose, every settings change your applications team walks a customer through, every quote your sales team builds, is knowledge that should be instantly retrievable and should compound over time. That is what we are building.


The approach

Start internal, prove it, then expand.

We deliver in stages, so each one is signed off and working before the next begins. This proposal prices Stages 1 and 2. The future stages are mapped at the end.

Now · Stage 1

Service Engineering KB

Searchable brain plus guided diagnostics for your engineers.

Next · Stage 2

Sales & Applications KB

Same engine, their content and access.

Future · Stages 3 to 5

Customer & growth

Portal, AI sales and quoting, then voice and bookings.

What we are building now

Three pieces, one engine.

STAGE 1

Service Engineering knowledge base

A searchable AI brain for your engineers and support staff, deployed where you choose.

  • Ingests service docs, bulletins, IFUs, knowledge bases, Sage Evolution records and emails. PDF, Word, text, email.
  • Reads images: attach a photo of a fault, error screen or component and ask about it.
  • Answers with the source documents cited, so engineers can verify every answer.
  • Learns from feedback: mark answers good or bad, and future answers improve.
  • Add knowledge by upload or paste, no technical skill needed.
  • Scales across Bone Density, Cardiology, X-Ray and more.
STAGE 1b

Guided Diagnostics & parts intelligence

The part of your scope that goes beyond a standard knowledge base. This is where your history pays off.

  • Step-by-step troubleshooting with the test procedures needed for further diagnosis.
  • Parts and repair recommendations ranked by fault probability, learned from your historical service records.
  • Clear instructions to execute the repair once the likely cause is found.
  • Scoped precisely after a short review of your service history. Recommended once Stage 1 is stable.
STAGE 2

Sales & Applications knowledge base

The same engine, pointed at your sales and applications team, with their own content and access.

  • IFUs, applications docs, quick reference guides, knowledge bases and Sage Evolution records.
  • Step-by-step guides to fix common problems, change settings, perform software analysis.
  • Same image analysis, source citing and feedback learn-loop.
  • Reuses the Stage 1 engine, which keeps its cost low.

How the AI handles your data

You set the privacy posture.

Your knowledge is proprietary and some of it is sensitive. This choice is separate from how the platform is hosted.

OptionWhat it meansBest for
Standard OpenAI Your data is never used to train OpenAI's models, and is only briefly retained for security monitoring before deletion. Lowest cost. Getting moving quickly at the lowest price.
OpenAI Zero Data Retention
RECOMMENDED
Nothing is retained by OpenAI at any point. Requires a committed annual AI spend of approximately $35,000, billed across 12 months. A medical business that wants strong privacy without running its own AI hardware.
JCR-hosted Australian AI A private AI model running in JCR's Australian data centre. Your data stays in Australia and never touches OpenAI. No hardware for you to buy. $1,500/mo exclusive use, or from $500/mo shared, depending on volume. Strong data sovereignty without a large AI commitment or your own hardware.
Fully local AI models The AI runs on dedicated hardware in your environment. Your data never leaves your premises at all. Maximum privacy, higher hardware cost. Maximum control. A natural fit with on-premise or self-developed hosting.

We recommend starting on OpenAI Zero Data Retention given the medical context, and keeping the fully local path open as you scale.

Three ways to deploy it

Same capability. Your call on hosting and ownership.

All three deliver the same Stage 1 and Stage 2 capability. Figures are indicative, ex GST, and confirmed after a short discovery.

Recommended to start
OPTION 1

Cloud-hosted Brain

$299 / mo
$1,500 setup

Amplor runs on JCR's managed Australian cloud. Access by secure login or your VPN. Fastest to stand up, lowest cost, fully managed by JCR.

  • You subscribe, JCR runs everything.
  • Standard, ZDR, or JCR-hosted Australian AI.
  • Live in 2 to 3 weeks.
OPTION 2

On-premise, JCR-maintained

$899 / mo
$5,500 setup

Deployed onto InMed's own VMs, exactly as your scope describes. Your data stays on your network, reached over your VPN. JCR maintains it remotely.

  • Updates, monitoring, backups, support included.
  • Standard, ZDR or fully local AI.
  • Live in 4 to 6 weeks.
OPTION 3

Fully self-developed, owned

from $45,000
build · then $900/mo support

A purpose-built platform developed for InMed, owned by InMed, running on your infrastructure. Works with any AI posture, standard through to fully local.

  • You own it outright, no lock-in.
  • Any AI posture: standard, ZDR, JCR AU or local.
  • Build time 6 to 12 weeks.
 Option 1 · CloudOption 2 · On-premiseOption 3 · Self-developed
Platform setup$1,500$5,500from $45,000
Platform monthly$299$899 incl. maintenance$900 support
Sage Evolution integration$3,900 once-off + $99/mo$3,900 once-off + $99/moincluded in build
Where your data livesJCR Sydney cloudYour own VMsYour own VMs
AI (monthly, on top)Standard ~$398, JCR AU from $500, ZDR ~$2,917Standard ~$398, JCR AU from $500, ZDR ~$2,917, or localStandard ~$398, JCR AU from $500, ZDR ~$2,917, or local
OwnershipJCR platform, you subscribeJCR platform, you hostYou own it outright
Time to first valueFastest (2 to 3 weeks)Medium (4 to 6 weeks)6 to 12 weeks
Best forQuick start, prove value fastYour stated VM preferenceTotal control, no lock-in

Your investment

Indicative, ex GST.

The example below is Option 1 (Cloud) with JCR-hosted Australian AI, our most cost-effective option that still keeps your data onshore. It covers Stage 1 and Stage 2.

ItemOne-timeMonthly
Amplor Brain platform
covers Stage 1 and Stage 2
$1,500$299
Sage Evolution integration
once-off
$3,900$99
AI processing
JCR-hosted Australian AI, onshore
no commitmentfrom $500
Indicative total$5,400from ~$898 / mo
Stage 1b Guided Diagnostics
optional, once stable
from $2,500included in platform

The example above uses JCR-hosted Australian AI at $500/mo shared, up to $1,500/mo exclusive. Two alternatives: standard OpenAI is the lowest cost at roughly $398/mo plus usage, with a short security-retention window; OpenAI Zero Data Retention adds ~$2,917/mo (~$35,000/yr committed) so nothing is retained by OpenAI at all, taking the monthly to about $3,315. On-premise and self-developed are priced per the table above; local AI hardware is specced and quoted separately. Existing JCR managed-services customers receive 15% off setup fees.

Timeline

Indicative, recommended cloud path.

PhaseDurationWhat happens
Discovery & setupWeek 1Confirm document volume, user numbers, AI posture and hosting. Stand up the environment.
Stage 1 buildWeeks 2 to 3Ingest service docs, bulletins, IFUs and emails. Image analysis, source citing, learn-loop live. Sage Evolution read integration.
Stage 1 tuningWeek 4Engineers use it, mark answers good or bad, we tune.
Stage 2Weeks 5 to 6Load the sales and applications corpus, configure the second audience and access.
Stage 1b (optional)Scoped separatelyGuided diagnostics and parts intelligence, after reviewing historical service records.

Your existing processes keep running in parallel throughout. Nothing is switched off.


Where this goes next

Your future list is already built.

Everything beyond the internal stages maps directly onto capabilities Amplor already has. When you are ready, we switch them on, stage by stage.

Stage 3
Customer self-service portalIFUs, troubleshooting and video cross-reference behind secure web logins. Amplor Web Chat and Brain in an authenticated portal on your website.
Stage 4
24/7 AI sales assistant and quotingProduct suggestions, drafted email responses your team approves before sending, and quotes generated from your Sage Evolution pricelists. The system learns from how you have responded to customers before. Amplor Inbox, Approvals and CRM.
Stage 5
Social, bookings, voice and email assistantsSocial media drafting, service scheduling and bookings, a virtual receptionist, and email assistants that draft replies for staff. Amplor Growth Engine, Bookings, Voice and Inbox, all shipped modules today.
To get started

A few decisions, a few handovers.

Decisions
  • Choose a hosting option: Cloud, On-premise, or Self-developed.
  • Choose an AI data posture: Standard, Zero Data Retention, or Fully local.
  • Confirm the first business units to load.
To finalise pricing
  • Rough document count, and how much is scanned PDF versus native text.
  • Internal user numbers across engineers and sales / applications staff.
Access we will need
  • Sage Evolution access, by API or direct database read.
  • A sample of historical service records, to scope Stage 1b.
  • A starting set of docs, bulletins and IFUs to load first.
Next step

The best next step is a quick chat.

We can run through it together, then firm up the final figures from your document volume and user numbers, and have a build plan ready within a day.

Reply to Joran →