Single agent · Specialty & small operators

One named agent. Fourteen days.
One workflow, owned outright.

For solo specialty subs, sub-$10M custom builders, and operators who want to validate one workflow before committing to a Fleet. $7,500 fixed-fee. Deployed in 14 days. Convertible to Fleet at any time, with credit applied.

Instant Agent

$7,500fixed-fee

14 days · 1 named agent · convertible to Fleet

One archetype, parameterized to your stack
Eval suite + production guardrails
Human-in-loop checkpoints enforced
30-day post-deploy support
Per-tenant data isolation
Convertible to Fleet — credit applied

Deploy an Agent → Book a Call First

What happens in 14 days

Day-by-day. Half the timeline of a Fleet. None of the corners cut.

Same eval rigor as a Fleet. Same per-tenant isolation. Same HITL gates. Just one agent, one workflow, one owner. Faster deployment because the scope is bounded.

Day 0 · Kickoff

Contract signed, archetype chosen

SOW signed. 50% deposit collected. Agent archetype selected from the 25-archetype library — Schedule Sentinel, Site Voice, Patient Recall, Doc Capture, etc. Stack inventory completed (10 minutes).

Day 1–3 · Integration

Read/write access provisioned to the target system

Read-only access by default; write access only on the target system the agent owns (e.g., Procore for Site Connect; Twilio for Site Voice; Athenahealth for Patient Recall). Per-tenant boundary enforced.

Day 4–7 · Configuration

Vertical schema + customer config layered onto the archetype

The archetype is the base. The vertical schema (Construction RFI flows vs. Healthcare patient-recall flows) is the overlay. Your customer config — thresholds, escalation paths, voice tone, brand language — is the third layer. All three locked before testing begins.

Day 8–10 · Eval & guardrails

Pre-deployment regression and quality evals run

Three categories: regression (does it still work the way the archetype is supposed to), quality (does the output meet the bar), safety (no hallucinations, no PII leaks, no scope creep). HITL gates configured for irreversible actions.

Day 11–13 · Shadow deployment

Agent runs in shadow mode against real production traffic

The agent observes real workflow events without acting on them. Outputs are reviewed by the customer's lead operator. Calibration loop: tighten thresholds, adjust voice, refine HITL triggers.

Day 14 · Go-live

Agent moves to production. 30-day post-deploy support begins.

Final invoice (50% balance) cleared. Agent live with full guardrails. 30-day post-deploy support window — direct Slack or email line to a Solutions Architect for tuning, eval refinement, or scope adjustments.

What customers actually deploy · three examples

Three Instant Agents that paid for themselves in the first 90 days.

Sample engagements across three of the five buckets. Each is a single agent, single workflow, focused recovery. Names anonymized; numbers documented under NDA but consistent with industry benchmarks.

CONSTRUCTION · CUSTOM BUILDER

Site Voice for after-hours owner calls

→ Site Voice

$8M custom builder. Owner calls coming in at 6–9 PM about finish changes were missed roughly 40% of the time. Site Voice answers, captures the change request structurally, routes to the PM by 7 AM the next day.

~$48K/yr in captured change orders

HEALTHCARE · DENTAL PRACTICE

Patient Recall for hygiene appointment lapses

→ Patient Recall

3-location dental practice. ~22% of patients past due on hygiene recall. Recall agent re-engages dormant patients via SMS + email + voice fallback, books appointments directly into Dentrix.

~$62K/yr in recovered hygiene revenue

REAL ESTATE · BROKERAGE

Schedule Sentinel for showing coordination

→ Schedule Sentinel

22-agent residential brokerage. Showing requests coming in across multiple agents fell through cracks. Schedule Sentinel coordinates, confirms, reschedules, escalates conflicts.

~$35K/yr in saved agent hours + retained leads

Fit qualification

Who Instant Agent is (and isn't) for.

Instant Agent is the right entry product for some businesses and the wrong one for others. Use the qualifier below before signing — or run a $9,500 DNA Scan first if you're unsure which agents you need.

✓ Fit for Instant Agent

Solo specialty sub or sub-$10M operator
One workflow you can name precisely
Recoverable revenue $30K–$100K/year on this single workflow
You already know which agent you need
Want to validate before committing to a Fleet
Stack is small enough to integrate one system cleanly

✗ Run a DNA Scan instead

Multiple workflows leaking — you don't know which to fix first
Recoverable revenue suspected above $250K/year
Operations span 3+ systems with handoffs between
$50M+ revenue with operational complexity
Healthcare PHI handling required
Want a recoverable-revenue dossier before committing to anything

Upgrade path · convert to Fleet at any time

Start with one. Add the Fleet when ready.

Most Instant Agent customers convert to Fleet within 6–12 months. The original $7,500 you paid for the Instant Agent applies as a credit against the first months of Fleet billing. No re-engineering — your Instant Agent becomes one of the 3–5 agents in your Fleet.

Month 0

Instant Agent live

$7,500

One named agent in production. 30-day post-deploy support. Recovering against one workflow.

→

Month 6–12

Fleet conversion

– $7,500 credit

Convert to Fleet ($4,500+/mo). The original $7,500 credits 100% against early Fleet billing. Add 2–4 more named agents from the same archetype library.

Instant Agent FAQ

Common questions, honestly answered.

Can I pick any of the 25 archetypes for my Instant Agent?

Yes — within reason. The 25 archetypes are: 5 universal (Intake, Compass, Guardian, Pipeline, Helix Memory), 5 operations (Site Voice, Schedule Sentinel, Doc Capture, Billing & AR, Reporting), and 15 vertical-specific add-ons (RFI Loop, Change Order Capture, Patient Recall, etc.). Helix Memory and Compass require a Fleet because they only deliver value when 2+ other agents are running. Everything else is available as an Instant Agent.

What if I don't know which agent I need?

Run an Operational DNA Scan ($9,500, 21 days). The Scan tells you exactly which agent (or which 3–5 agents) to deploy first, ranked by recoverable revenue. Most Scan customers convert to Fleet, not Instant Agent — but if your Scan dossier identifies a single high-value workflow, Instant Agent is the right choice.

How is this different from just using Zapier or Make?

Zapier moves data between systems on rule-based triggers. An Instant Agent owns an outcome — Patient Recall doesn't just send reminders; it adapts the message based on patient history, escalates to voice when SMS fails, books appointments directly, and logs the result. The agent has goals and judgment under HITL gates. Zapier has triggers and actions. If your workflow is rule-based, use Zapier — it's 10× cheaper. If your workflow needs decisions, you need an agent.

Does Instant Agent get the same eval suite as Fleet?

Yes. Same regression eval, same quality eval, same safety eval. The eval suite is per-archetype, not per-customer — so an Instant Agent gets the full eval rigor of every Fleet that's deployed the same archetype before it. The only Fleet-exclusive item is continuous drift monitoring; Instant Agent gets 30-day post-deploy monitoring, then the customer either converts to Fleet (with continuous monitoring) or operates the agent under their own observability.

What happens after the 30-day post-deploy support window?

Three options. One: convert to Fleet ($4,500+/mo) — gets you continuous monitoring, eval refresh, and 2–4 more agents. Two: month-to-month maintenance ($750/mo) — agent stays under our observability and gets monthly eval runs, but no new agents added. Three: hand off — we deliver the agent code, configuration, and eval suite; you operate independently. Most customers pick option 1 or 2.

Is Instant Agent BAA-eligible for healthcare workloads?

Yes. PHI workloads route only to BAA-eligible LLM tiers (Anthropic Enterprise, OpenAI Enterprise, Google Cloud HIPAA-eligible) and BAA-capable sub-processors. See the BAA process on the compliance page. The Healthcare Instant Agent (e.g., Patient Recall, Prior Authorization) is fully BAA-deployed.

Can I cancel mid-deployment?

If you cancel before day 7, you receive a 50% refund of the $7,500 (we keep the deposit and prorated work). If you cancel after day 7, the deposit is non-refundable but you keep the agent code and configuration delivered to date. We never had a customer cancel mid-deployment, but the policy is published so you know going in.

What's the minimum revenue threshold to make this worthwhile?

Roughly $30,000/year in recoverable revenue on the single workflow. Below that, Zapier or a part-time VA is more cost-effective. Above that, Instant Agent's $7,500 fixed cost amortizes inside 90 days. If you can't name $30K of recoverable revenue, run a DNA Scan first — it tells you whether the math works.

One named agent. Fourteen days.One workflow, owned outright.