Illustrative scene, not actual participants or clients
For your business
A bounded assistant workflow with allowed tools, approval points and trial criteria.
For you as a leader
Clarity about which follow-ups an assistant could prepare and where you or your team must approve.
REAL OUTPUT
What your team may take away
Illustrative sample output, not an actual client documentSample-data assistant workflow from intake and sources to draft and human approval
Illustrative sample output, not an actual client documentAllowed-data and tool checklist with least-privilege access, review points and error response
Illustrative sample output, not an actual client documentFour-week pilot brief with baseline, target, owner, tool costs and go-or-stop criteria
Know who you will work with and what is included
We match facilitators and specialists to the challenge and learners. The proposal identifies the delivery team, roles, activities and scope before agreement.
The four-week pilot plan is an output for your team to use; four weeks of follow-up services are not automatically included.
Final content, deliverables and duration are agreed after we understand your challenge.
Hermes Agent is a Nous Research project. This is a practical work-design workshop, not a developer-endorsed certification. Nous Research ↗
Indicative budget
1 day / 6 hours: THB 59,000 · 2 days / 12 hours: THB 115,000 · Up to 15 participants
One course per cohort · Thai-language delivery on weekdays at your organisation
Included
Facilitation, digital materials, templates and review during workshop hours.
What your team takes away
Course outputs and a pilot or practice plan for your team to use, within the course scope.
After the workshop
Post-workshop follow-up and production deployment are not included. Additional work is scoped separately.
VAT and final price
VAT, if applicable, and additional costs are assessed separately and confirmed in the quote before agreement.
What preparation is included, and how do one and two days differ?
One online briefing of up to 45 minutes; activities adapted from the existing course for one main challenge; facilitator, coordination, digital materials, templates and review during workshop hours.
AI packages include one workshop assistant.
1 day: learn, practise and draft outputs. 2 days: work more deeply on the same course, with feedback and revisions around your team’s challenge.
Which costs are assessed separately?
Venue, catering, travel, accommodation, VAT if applicable, software, English-language delivery, weekends, additional participants and wholly new course design are quoted separately before agreement.
What if we need consulting or help after the workshop?
Production deployment and post-workshop follow-up services are not included. Pilot and practice plans are outputs for your team to use afterwards.
Organisational transformation consulting, AI adoption and bespoke process improvement are scoped and quoted separately.
Budgets reflect the scope above. Final pricing and the delivery team are confirmed in the proposal before agreement.
Track completed follow-up steps, errors, review time and cases that need escalation.
Include workshop fees, learning time, tools and human review. The value of released capacity is not automatically a cash saving.
Calculation example — hypothetical figures
Illustrative numbers: 30 follow-ups a week
Planning assumptions, not client results or a promise. Replace them with your organisation's baseline before deciding whether to use the tool.
Time to find status and draft
60% less
30 items × five minutes = 150 minutes a week. The trial target is two minutes each, or 60 minutes a week, including human review.
Value of capacity released
THB 28,800/year
Ninety minutes, or 1.5 hours a week × an assumed THB 400 loaded labour cost an hour × 48 weeks. This is capacity for other work, not cash savings or guaranteed profit.
Overdue follow-ups
50% fewer
An illustrative eight a month could become no more than four, while recording why any item still slips.
How to verify it: capture at least two weeks of baseline data, run a four-week low-risk trial, then count human review, corrections, overdue items and tool costs before granting wider access.
Questions before deciding
How much time is needed?
1 day: select a suitable task, practise on sample data and set permission boundaries. 2 days: workshop your team's workflow, approval points and measurable pilot plan.
What prior knowledge does the team need?
We discuss participants’ experience, roles and challenge before agreeing the activities. Bring a laptop and a follow-up task you can describe without confidential details. Accounts, tools and organisational permissions are agreed before live use.
Do we need company data?
Begin with sample data or examples you can share. Live data, permissions and tools must be agreed under your organisation’s requirements.
What determines the cost?
Fees depend on the challenge, participants, duration and outputs. Also consider tool costs and your team’s implementation time. System deployment or managed operations are not automatically included.
Who follows through after the workshop?
The four-week pilot plan is an output for your team to use; four weeks of follow-up services are not automatically included. Your organisation assigns owners to test and measure the work. Additional services require a separate agreement.
Compare this with your team’s situation
After a Wednesday meeting, a team manager has a dozen loose ends. A customer awaits a document, Sales needs a warehouse answer, and leadership wants a summary by three. Some items are in email, some in chat, and some in a notebook. When someone asks what may slip, the team starts searching again. Small follow-ups have eaten the time needed to solve the real issue.
Illustrative scenario, not an actual client case
Explore the approach to this scenario
Nous Research's Hermes Agent can use tools and memory in multi-step work. That does not mean giving it unrestricted access. We begin with a bounded sample-data workflow: receive an item, check permitted sources, draft a summary with uncertainties, then stop for a person to review and approve any consequential message or change.
Imagine 30 follow-ups a week, each taking five minutes to locate and draft: 2.5 hours in total. In class, we test whether an assistant could prepare enough context to bring that to two minutes per item including human review. We use fictional information, not connected production accounts, and do not let an agent silently send messages or edit records.
IN THE ROOM
What we practise together
Participants play manager, data owner and reviewer. They test stale information, missing access, uncertain answers and out-of-scope requests to decide where the assistant must stop and ask a person.
Illustrative scene, not actual participants or clients
01
Choose work an assistant may prepare: separate finding, drafting and reminders from human decisions; measure volume, current effort and the cost of mistakes.
Illustrative scene, not actual participants or clients
02
Practise a multi-step task with sample data: group open items, draft a sourced update, then inspect how stale memory or missing context can distort an answer.
Illustrative scene, not actual participants or clients
03
Set the guardrails: specify allowed tools and data, least-privilege access, approvals, an audit trail and a stop point for unclear or out-of-scope requests.
Illustrative scene, not actual participants or clients
04
Plan a four-week trial: name an owner, permitted tasks, baseline effort and missed follow-ups, review method, tool costs and stop criteria before expansion.
ThinkToFinishTAILORFIT LEARNING
COURSE BRIEF
Course brief to share
A starting summary for your team and approvers
A Work Assistant with Hermes Agent
Let an assistant prepare follow-ups while your team keeps control
Who it is for
Team managers, executive assistants, operations and supporting IT teams
What to bring
Bring a laptop and a follow-up task you can describe without confidential details. Accounts, tools and organisational permissions are agreed before live use.
Indicative duration
1 day: select a suitable task, practise on sample data and set permission boundaries.
2 days: workshop your team's workflow, approval points and measurable pilot plan.
Possible outputs
Sample-data assistant workflow from intake and sources to draft and human approval
Allowed-data and tool checklist with least-privilege access, review points and error response
Four-week pilot brief with baseline, target, owner, tool costs and go-or-stop criteria
Indicative budget
1 day / 6 hours: THB 59,000 · 2 days / 12 hours: THB 115,000 · Up to 15 participants
One course per cohort · Thai-language delivery on weekdays at your organisation
One online briefing of up to 45 minutes; activities adapted from the existing course for one main challenge; facilitator, coordination, digital materials, templates and review during workshop hours.
AI packages include one workshop assistant.
Venue, catering, travel, accommodation, VAT if applicable, software, English-language delivery, weekends, additional participants and wholly new course design are quoted separately before agreement.
Production deployment and post-workshop follow-up services are not included. Pilot and practice plans are outputs for your team to use afterwards.
Budgets reflect the scope above. Final pricing and the delivery team are confirmed in the proposal before agreement.
Final content, deliverables and duration are agreed after we understand your challenge.
