Golden Horizons
CAPABILITY · OPS & BACK-OFFICE
Admin Assistant
Fully autonomous internal assistant handling scheduling, docs, and staff requests.
$7,500–$12,000 build · $2,500–5,000/mo
Talk to us about a Admin Assistant build →What it does
Handles scheduling, document drafting, internal Q&A, and routine staff requests via Slack or email. Routes tasks to the right person, follows up on blockers, and keeps the owner out of the day-to-day. Self-hosted path available for data-sensitive clients.
The math on a full-time EA rarely works for a seven- or eight-person firm. A dedicated executive assistant runs $55,000–$75,000 in salary before benefits, PTO, and the six weeks it takes to onboard them to your calendar preferences, your CRM, your travel vendors, and the way you like agendas formatted. Most owner-operators at that stage are doing the EA work themselves — fielding scheduling threads that take four back-and-forth emails to resolve, blocking an hour before every client call to pull background they already have somewhere in their inbox, drafting the same meeting recap email with the same five bullet points, and chasing the contact they forgot to log in the CRM three days after the conversation.
The Admin Assistant build is scoped to the tasks that consume time without requiring judgment. Scheduling runs through your calendar with awareness of your buffer rules and travel time. Research briefs get pulled from public sources and your own notes before a call, formatted the way you actually use them. Short-form writing — follow-up emails, meeting agendas, brief status summaries — gets drafted and queued for your one-touch approval rather than written from scratch. CRM records get updated from meeting notes without manual data entry. Travel and logistics get coordinated within the parameters you set. Reminders and follow-up nudges fire on the schedule you define.
What it doesn't do is handle calls that require your professional judgment, client-relationship decisions, or anything where being wrong has real consequences. Those escalate to you — cleanly, with the relevant context attached — so you're making the call, not hunting for the background. The system lives where you already work: Slack, email, or SMS, depending on your preference. It doesn't require a new app, a new login, or a new habit. You message it the same way you'd message a person, and it handles the task or asks the one clarifying question it needs before moving.
The build takes three to four weeks. The first week is scoping — we map exactly which recurring tasks make sense to hand off and which ones stay with you. Weeks two and three are build and integration. Week four is tuning: we run it alongside your real workload, watch where it hesitates or over-escalates, and tighten the routing rules until the signal-to-noise ratio is right. After that, the retainer covers ongoing tuning as your workflow shifts and integration maintenance when the tools it touches push breaking updates.
Use cases
- A solo consultant books twelve client calls a month. Scheduling threads eat forty minutes per booking. After the build, a single message to the assistant resolves each booking in one exchange — the client gets a link, the consultant gets a confirmed event with pre-call prep notes already attached.
- A real estate agent runs fifteen active showings per week. The assistant queues the day's showing schedule each morning, texts confirmation reminders to buyers, and logs post-showing notes from a voice memo into the CRM — without the agent touching a keyboard between appointments.
- A managing partner at a boutique law firm spends ninety minutes every Sunday night on the week's calendar and agenda prep. The assistant handles both — reviewing the calendar, pulling relevant case notes or client background for each meeting, and drafting a Monday briefing the partner edits in ten.
- A government contractor proposal manager tracks deadlines across six active bids. The assistant monitors due-date fields in the project tracker and sends a daily digest of upcoming milestones, flags any deadline within five business days, and drafts the internal reminder email for the team — no.
- An insurance broker follows up with prospects across a sixty-day pipeline. The assistant monitors last-contact dates in the CRM and drafts personalized follow-up emails for any prospect past the defined touch interval, queuing them for one-click send rather than requiring the broker to write each.
- An accounting firm manager handles inbound staff questions about PTO balances, reimbursement policy, and internal process documentation. The assistant answers from a connected knowledge base, handles the eighty percent of questions it can resolve completely, and escalates the rest with context.
What’s included
- Fixed scope with written acceptance criteria before any build starts
- Customization layer for your brand voice and business rules
- Clean handover with documented runbook and live training
- Monthly ROI report for three months post-delivery
- Source code delivered to your GitHub on handover
What’s NOT included
- Third-party API subscription costs (billed to your accounts)
- Data migration from legacy systems
- Ongoing infrastructure costs after handover
Retainer
Monthly retainer covers monitoring, prompt tuning, config refinement, and minor integration additions. Range: $2,500–5,000/mo.
How clients use this
Fixed-scope build with clean handover, then an optional monthly retainer covering maintenance, monitoring, and minor changes. Most clients move to retainer within 60 days of delivery.
Part of
Used in: Law Firms , real-estate-agents
Questions Admin Assistant clients ask
What access does the assistant need to my calendar, email, and CRM?
We use scoped OAuth credentials — not admin access. For Google Workspace, that means calendar read/write limited to your account and a Gmail send permission scoped to drafts and sending from your address only. For the CRM, we use a dedicated API user or integration token with write access limited to the specific objects the build touches: contact records, activity logs, and task fields. We document every permission on paper before anything is connected, and you approve the scope before credentials are issued. If you use a self-hosted CRM or have data-residency requirements, we can deploy the integration layer inside your own environment so data doesn't leave your perimeter. We don't take blanket admin credentials, and we don't retain access to accounts after the engagement closes — credentials are rotated or revoked at offboarding.
Which tasks will the assistant handle autonomously versus escalate to me?
The routing rules are defined during the scoping week and documented in plain language — not buried in a system prompt you can't read. Tasks that run autonomously are the ones with deterministic outcomes: scheduling against known calendar rules, sending pre-approved template emails, updating CRM fields from structured inputs, pulling research from defined sources, firing reminders on a set schedule. Tasks that escalate are the ones where being wrong costs something: any outbound message that isn't templated, any decision involving a client relationship, any situation the assistant hasn't seen before and can't confidently classify. Escalations arrive in Slack or email with the relevant context already attached — what the task was, what the assistant tried to do, and why it stopped. You make the call in one message, and it executes from there. The escalation threshold is tunable: if it's escalating too much, we tighten the rules. If it's running too autonomously for your comfort, we widen the escalation criteria.
How long does it take before the assistant is actually useful?
The build takes three to four weeks. By the end of week four, it's handling the tasks we scoped without hand-holding. That said, the first two weeks after go-live are the highest-value tuning window — we watch the escalation log, identify patterns where the assistant is hesitating unnecessarily or missing context it should have, and adjust the routing rules. Most builds reach a stable operating state within thirty days of launch. 'Stable' means the escalation rate is low, the tasks it owns are running without intervention, and you've stopped second-guessing it. After that, the main tuning events are when your workflow changes materially — a new hire, a new tool in the stack, a shift in how you structure your week — or when an integrated platform pushes a breaking API update. The retainer covers both.
Where does the assistant live — do I need to install a new app?
No new app, no new login. The assistant runs inside Slack, email, or SMS — whichever channel you already use. Most owner-operators default to Slack because it's already where short requests live. If you don't use Slack, we wire it to a dedicated email address or a phone number, and you interact with it the same way you'd message a person. The backend infrastructure runs on your existing cloud environment or ours, depending on your preference. There's nothing to install on your end. The only visible change is a new contact in your communication tool that handles admin tasks instead of ignoring them.
How is this different from just using ChatGPT or an AI tool I already have?
The gap is integration and memory. A general AI tool answers questions in a conversation window. It doesn't have access to your calendar, your CRM, your email history, or your specific business rules. Every session starts from zero — you re-explain your preferences, re-paste the context, re-describe the output format you want. The Admin Assistant build is wired directly into the tools your business already runs on, knows your calendar rules and contact records without being told, operates on a schedule without you initiating it, and applies the routing logic we tuned to your specific workflow. Golden Horizons builds the integration layer, the memory scaffolding, and the escalation logic that connects a capable AI model to your actual operating environment — so it works like a hire, not like a search engine.