Gonzo Digital / fixed-price systems for routine work

Someone asked. Nobody answered.

By the time you reply, they have booked with whoever did. We build the system that answers in about a minute (a design target, not measured live) and hands you only what needs a person.

From $1,500 setup and $500 a month, fixed. US dollars; local currency and tax are confirmed on your proposal.

Get the free auditSee it run

Free audit, no accounts, we change nothing.

Or ask a person: aimarketing@gonzo.co.in +91 7390990227 WhatsApp

  • Always tells you it is AI

Gonzo Digital is an India-based company that sells nine fixed-price systems. Each takes over one routine job. Setup is $1,500 to $5,000 once, then $500 to $2,500 a month, in US dollars. Live in 7 to 21 business days. Every message a system sends on its own says it is from an AI.

9

fixed-price systems

7–21

business days from a signed scope and the deposit to live

60 s

first-reply target on Lead Response, day or night. A design target, not measured live.

100%

of your ICP-fit leads that have a consent record get contacted. A count in the weekly report, not a claim about lead quality.

Check us before you talk to anyone.

Nobody is live yet, so there are no client results. Here is what you can check instead.

Honest status: nobody is live yet. First clients get a fixed scope and a written handover. Trust and method.

Start here

Pick the task you hate. Start with one of these two.

Our two most built systems. Both start with people who already wrote to you.

Lead Response Engine

Inquiries sit unanswered for hours. One of the two most built: email and WhatsApp are tested in a sandbox, never run live. The web form and chat adapters are not built yet.

$2,000setup, once
$600per month

Live in 10 business days from a signed scope and the deposit.

See what you get →

Dormant List Revival

Past customers nobody writes to. One of the two most built: email and WhatsApp are tested in a sandbox, never run live. Going live needs a sending domain and a consent record per person.

$1,500setup, once
$500per month

Live in 7 business days from a signed scope and the deposit.

See what you get →

Fixed scope, fixed price, in US dollars. The India list is in INR. Local currency and tax are confirmed on your proposal.

All nine systems and prices

What is built today, and the India list in INR.

Models used: Hermes and DeepSeek, product by product. Details

Connections built so far. All tested in a sandbox; none has run live.

  • Email
  • WhatsApp Business API
  • Web chat
  • Google Calendar
  • Microsoft 365
  • Stripe payment links
  • Your CRM
  • Your website

Fixed scope, fixed price, named deliverables. Each system is run by a named agent, shown on every card.

USD is the list price for businesses outside India.

01General lane

Lead Response Engine

$2,000setup
$600per month

Live in 10 business days · Year one $9,200

Today: Sandbox only. Web form and web chat adapters not built.

  • Inbound capture from WhatsApp and email into one queue (the web-form and web-chat adapters are not built yet)
  • First reply aimed at 60 seconds, 24/7, in the visitor's language (a design target, not measured live)
  • Qualifying questions asked automatically before a human is looped in
02General lane

Dormant List Revival

$1,500setup
$500per month

Live in 7 business days · Year one $7,500

Today: Sandbox only. Needs a sending domain and a stored consent record for every person.

  • Segmentation of the existing customer list by recency and value
  • A written win-back sequence (WhatsApp + email), staged, not blasted
  • Automated send with a human-set daily cap and an honored opt-out
03General lane

Client Onboarding Pipeline

$2,500setup
$700per month

Live in 12 business days · Year one $10,900

Today: Sandbox only. Calendar never run live; state store not built.

  • Signed-contract trigger starts the pipeline automatically
  • Intake form sent, chased, and logged without a human remembering to
  • Kickoff scheduling handled end to end
04General lane

AI Receptionist

$1,800setup
$800per month

Live in 14 business days · Year one $11,400

Today: Chat today. Voice needs a telephony gateway that is not built. Sandbox only.

  • 24/7 chat answering today (bookings, FAQs and routing); phone calls need a telephony gateway that is not chosen or built yet
  • Calendar-integrated booking with double-booking prevention
  • Escalation to a human for anything outside its script
05Finance lane

Quote-to-Invoice Automation

$2,000setup
$600per month

Live in 10 business days · Year one $9,200

Today: Sandbox only. Stripe never run live; accounting package not connected.

  • Quote generation from a standard input form, sent within minutes
  • Automatic invoice on acceptance, no manual re-entry
  • Staged payment reminders that stop the moment payment lands
06General lane

AI Outreach & Lead Qualification Engine

$3,000setup
$1,200per month

Live in 14 business days · Year one $17,400

Today: Sandbox only. Can write first, to business contacts only; cold email depends on the country.

  • ICP-fit lead scoring with the evidence behind every score, never a black-box percentage
  • AI-drafted outreach sequences that disclose they're AI on every first touch, on every channel, by design, not a setting you can turn off
  • Runs only on Hermes and DeepSeek (how personal data is handled is described on the trust page)
07General lane

E-commerce Order, Returns & Product Support Assistant

$1,800setup
$700per month

Live in 12 business days · Year one $10,200

Today: Sandbox only, against a simulated shop. No refund function.

  • Order-status lookup, answered instantly from your own order data
  • Return eligibility checked against your stated policy, with the exact rule cited every time, never a black-box yes/no
  • Technical spec Q&A that answers only from your product data on file, and says so the moment a question isn't covered
08General lane

Managed Automation & Infrastructure Retainer

$2,500setup
$2,500per month

Live in 7 business days · Year one $32,500

Today: Sandbox only. Not a 24/7 human helpdesk.

  • We run and maintain every Gonzo system deployed for you: monitored, fixed and tuned as your business changes, not left to rot
  • A named point of contact and a real response-time commitment (SLA), reported on monthly
  • Quarterly review: what's working, what to add next, what to cut
09General lane

Connected Operations Layer (MCP)

$5,000setup
$2,500per month

Live in 21 business days · Year one $35,000

Today: Least finished. Router never run live; most write connectors not built.

  • One AI operator connected across your whole stack (mailbox, calendar, CRM, documents and WhatsApp Business) through the Model Context Protocol, acting as a single surface instead of ten disconnected apps (today only a few read tools are built; most write connectors are not, and the router has never run live)
  • Every irreversible action (sending, paying, publishing, deleting) stays human-approved by construction. The agent drafts and proposes; you approve. This is the safety that makes 'connect everything' sane
  • WhatsApp handled through the official Business API within its real limits: own-account messaging and inbound replies, never the stranger-outreach automation that WhatsApp's policy does not allow and that can get accounts suspended

What happens next

  1. Free auditAbout five business days. A short written report on one page. We change nothing.
  2. Fixed-price scopeIf a system fits, you get its scope and price in writing. Nothing is paid on this site.
  3. Build7 to 21 business days from a signed scope and the deposit. The written handover comes before go-live.
Get the free auditSee it run

Free audit, no accounts, we change nothing.

Limits

What this does not do.

  • It does not pretend to be a person. Every first touch says it is an AI.
  • It does not move money out. No refunds, payouts or transfers. It passes the details to you.
  • It does not send cold WhatsApp or ring strangers. No bought or scraped lists.
  • It does not promise rankings, citations or results.

All eleven limits, in full
  • It does not send cold or broadcast WhatsApp, make cold calls or use bought or scraped lists. Most systems write only to people who wrote to you or agreed to hear from you. The Outreach Engine is the one that can write first: only to business contacts from public business data, an introduction or an inbound inquiry, each cleared by a consent check, with AI disclosed. Whether cold business email is allowed depends on the country.
  • It does not pretend to be a person. Every first touch says it is an AI, on every channel.
  • It does not move money out. No refunds, credit notes, write-offs, payouts, transfers or chargeback replies. Those functions do not exist in it. It passes the details to you.
  • It does not sign or trade. It cannot sign a contract for you, and it cannot place a trade.
  • It does not scrape. The growth agents read your own site only and obey robots.txt. No LinkedIn automation.
  • It does not post on social media for you, and it does not touch app-store listings.
  • It does not promise rankings, citations or results. Nobody controls where Google or an AI answer puts you.
  • It does not use an AI model without saying which. The models are Hermes and DeepSeek. The Outreach Engine runs only on them, two other systems list them in their stack, and for any other a person confirms in writing. How personal data is handled is on the trust page.
  • It does not ask for your passwords. You connect your accounts through each provider's own sign-in and can revoke access at any time. Where access tokens are held depends on the connection route and is confirmed in writing before go-live.
  • It does not replace your IT team. The retainer is not a 24/7 human helpdesk, and it does not cover on-prem hardware or IT we did not build.
  • It does not guess. If your data or your policy does not cover a question, it says so and passes it to a person.
  • It does not decide complaints or legal matters. It stops after one neutral reply and tells you.

The question everyone asks

What happens if we disappear.

  • You own it.The workflow, the accounts and the data are yours from day one.
  • You can revoke access.Accounts connect through each provider's own sign-in. Revoke it whenever you like.
  • You can carry on without us.You already hold the workflow, the accounts and the written documentation.

The full answer, and how the handover works
01

You own it.

The workflow, the accounts and the data are yours from day one, under your own provider accounts. Not ours.

02

The access is yours to revoke.

Accounts connect through each provider's own sign-in, and you can revoke access at any time. Where the access tokens are held depends on the connection route, and a person confirms that in writing before go-live.

03

You can carry on without us.

If we vanish, you already hold the workflow, the accounts and the written documentation. You can revoke our access at each provider and export your data without us.

Credentials and documentation are handed to you in writing before we mark a system live, not after. This is a gate in our own delivery process, not a promise. How delivery works.

Look inside

Open any of these. Everything is shown in full.

Watch a sandbox run.

Simulated customers, no real money. Flip the kill switch and watch it stop.

Each line is one action an agent took. Teal is routine work, orange is anything about money, white is a handoff to a person.

gonzo autopilot / sandbox Sandbox run, simulated data
Sent (sandbox)
339
Held in outbox
0
Came to you
7
Denied
2
Retries healed
51
Quarantined
9
Owner taps
5
Audit entries
812
Checks held
79/ 79
Audit chain
verified
Mon 09:00:00generalkernelBootAUTOPILOT_ENABLED starts OFF: every external send would go to the dry-run outbox. Sandbox mandate loaded.
Mon 09:00:00generalkernelSafety checkKill switch readable (halted=False). Circuit breaker armed per agent.
Mon 09:00:00generalkernelFlag flipAUTOPILOT_ENABLED flipped ON by shreyas. One recorded flip. Sandbox transports only.
Mon 09:00:01generalgonzo_retainer_watchdogProbehandled 5, sent 0 (probes taken 3)
Mon 09:01:02generalgonzo_lead_responseDead letteract:route_lead: gave up (permanent) after 3 attempt(s); dlq-0001
Mon 09:01:03generalgonzo_lead_responseDead letteract:route_lead: gave up (permanent) after 1 attempt(s); dlq-0002
Mon 09:01:04generalgonzo_lead_responseInquirieshandled 9, sent 9, handed off 1 (replied 8, kb answered 4, kb misses 1, handoff lane 1, finance desk handovers 1)
Mon 09:01:04handoffgonzo_lead_responseHandofflr-c09: handed off (complaint, human_request)
Ready
sandbox run, simulated data. Seed 7. Fault injection 0.15. This run and the one in the dashboard demo are two separate simulated runs, with different seeds and different businesses, so their numbers differ. In production this is the owner's own kill switch: every agent checks it before every action, and if it cannot be read, the agent does not act.
generalfinancehandoff

Money questions get answered. Money decisions stay yours.

Only inside limits you set. Anything above them comes to you.

Price and billing questions, objections, negotiation, quotes, invoices, payment links, reminders and reconciliation are handled by a finance desk, only inside limits you set. Anything above those limits comes to you. Three outcomes, replayed below on simulated conversations.

The mandate

Sandbox, illustrative

These limits are the owner's to set. The numbers shown are illustrative. Until they are set, everything in finance comes to you.

Maximum discount
10%
Price floor
85%of list
Installments, at most
3
Payment terms, at most
14days
Largest quote
$15,000
Ad spend per day
$50cap

Fingerprint 88722047c5007262… Every decision is logged with it. A mandate expires, and an expired one counts as unset.

A price question

Buyer (simulated)
How much does the AI receptionist cost per month?
GonzoAI
AI Receptionist: setup $1,800 (one time, on signing) and $800 a month (billed monthly, with the payment method confirmed in writing before you sign). The first year comes to $11,400, setup included. (This message was drafted by an AI assistant on behalf of Gonzo Digital regarding pricing and billing. Reply and a person will take it from here.)
Decision: answered from the price list; logged with mandate fingerprint 88722047c500
Handled inside the mandate.Nobody had to be asked. It is in the log with the mandate fingerprint.

On the desk

  • gonzo_finance_qaAnswered 5Held For Person 2Answers price and billing questions from the price list.
  • gonzo_objectionsHandled 10Escalated 0Handles "too expensive", "not now" and "we tried an agency".
  • gonzo_negotiatorCounters Inside Mandate 3Accepted 2Counters inside your limits. Sends the rest to you.
  • gonzo_quote_to_invoiceQuotes Sent 2Invoices Sent 2Issues quotes and invoices, and sends payment links.
  • gonzo_collectionsInvoices Enrolled 3Stopped On Payment 1Sends payment reminders. Stops when the payment lands.
  • gonzo_reconciliationEvents Received 2Payments Applied 1Matches payments to invoices and flags mismatches.
  • gonzo_adsLaunches 1Pauses 3Plans ad spend inside the cap you set.

Owner queue (sandbox)

  • Answer billing question9 in this run. A money question the desk could not answer from the price list, or an offer above its limits.Pending owner
  • Client target down1 in this run. A retainer client's system stopped answering its health checks.Pending owner
  • Dead letter unfixable3 in this run. A failed action the system could not replay by itself. Shown to the owner, never retried unseen.Pending owner
  • Finance voice handover1 in this run. A caller asked about money by voice. The desk has no voice channel, so a person writes back.Pending owner
  • Payment mismatch1 in this run. A payment that differs from its invoice. Not closed automatically.Pending owner
  • Price quote list1 in this run. A price answer that needed the owner before it could be sent.Pending owner
  • Refund request3 in this run. Refund asked for. These agents have no refund function: the details were taken, a holding reply was sent and a person decides.Pending owner
  • Return exception review1 in this run. A return outside the stated policy window, held for a person.Pending owner
  • Support escalation2 in this run. A support question outside the rules or the product data on file.Pending owner
  • Tax treatment needs owner1 in this run. A quote held because no tax rule is on file for that buyer.Pending owner
  • Answer billing question1 in this run. A money question the desk could not answer from the price list, or an offer above its limits.Approved by owner
  • Issue invoice1 in this run. An invoice above the mandate's value limit.Approved by owner
  • Issue quote1 in this run. A quote above the mandate's value limit.Approved by owner
  • Price quote list1 in this run. A price answer that needed the owner before it could be sent.Approved by owner
  • Revival campaign start1 in this run. The first-wave go-ahead for a win-back campaign (a demonstration of the owner's queue).Approved by owner
  • Handoff20 in this run. A complaint, legal wording, a VIP, a request for a person or an erasure request. The agent stopped after one neutral reply.Handed off

Never, at any level: refunds, credit notes, write-offs, payouts, transfers, chargeback replies, trades and signatures. There is no function in the agents for any of them. The agent can take the details, state the policy, send a holding reply and pass it to you. That is all.

One checkpoint. Every action goes through it.

Seven checks, three ways to stop it, a log you can test.

Nothing leaves the building without passing the same seven checks, in the same order. That is the whole design. It is what lets the routine work run without someone tapping approve on every message.

  1. Kill switch and breaker

    Every agent asks first. If it cannot read the switch, it does not act. Repeated failures trip a breaker that only a person can reset.

  2. Finance sort

    Money language, in or out, is routed to the finance desk. A routine agent that writes a price is blocked and rerouted.

  3. Consent and opt-out

    No consent record, no message. STOP is honored at once.

  4. Your limits

    Finance acts only inside the limits you set. Above them, it comes to you.

  5. AI disclosure

    The first touch says it is an AI, on every channel. There is no setting to turn that off.

  6. Outbound scan

    The finished draft is checked again before it leaves.

  7. Audit log

    Every action is written to a tamper-evident log.

The autonomy ladder

You choose how far up it goes. Pick a rung to see what it allows.

  1. Rung 0: Paused. Nothing runs. The kill switch is engaged and every agent checks it before every action.
  2. Rung 1: Dry run. The agents do the work but nothing leaves. Every message waits in an outbox you can read. This is where it starts.
  3. Rung 2: Routine work. Agents send routine messages on their own: replies, follow-ups, reminders. Consent, AI disclosure, opt-outs and the daily cap still apply. Anything about money goes to the finance desk.
  4. Rung 3: Finance inside your limits. The finance desk quotes, counters and invoices on its own, only inside the limits you set. Anything above them comes to you.
What each rung allows
RungPausedDry runRoutine workFinance inside your limits
Agents read, draft and scoreNoYesYesYes
Messages leave the buildingNoOutbox onlyRoutine onlyRoutine and finance
Quotes, invoices, payment linksNoDrafts onlyGoes to youInside your limits
Anything above your limitsNothing runsComes to youComes to youComes to you
Refunds, credit notes, payouts, trades, signaturesNo capabilityNo capabilityNo capabilityNo capability

The brakes

Three ways to stop it, all yours.

Kill switch

Stops every agent before its next action. If the switch cannot be read, agents do not act.

Circuit breaker

Trips after repeated failures. Only a person can reset it.

Pause Gonzo

One command pauses every agent at once.

  • Handoff, not argument. A complaint, a large or important account, or legal and compliance language: the agent stops after one neutral reply and tells you. It never argues or negotiates.
  • Deterministic first. Rules and templates decide what happens. The model only helps with wording. Email addresses, phone numbers and amounts are swapped for placeholders by pattern before any model sees the text, and names where they have been supplied, then put back afterwards. That lowers the risk; it is not proof that no personal data remains.
  • Caps. A daily message cap you set applies to every sending agent. The sandbox run used 200 a day.

A log that shows if anyone touches it

Each entry carries the fingerprint of the one before it. Change a single entry and every entry after it stops matching. The head fingerprint below is from the sandbox run.

Entries in the run
812
Chain
Verified
Checks held
79 of 79
Head fingerprint (SHA-256)
98689bdd49180faa9b11afa43a2a90503c7c8fc536c85b72f387c8b806b58eb0
Illustration built in your browser from the replayed events. Click a block to alter it.

Be found in Google, and in AI answers.

Three agents outside the nine. No ranking promised.

Three growth agents, kept inside limits. They read your own site, not other people's. They are not part of the nine fixed-price systems. Ask us about scope.

SEO / Google

Search in Google

Your own pages, found by the people who are already searching for what you do.

  • Reads your own site, and only your own site. It reads robots.txt first and obeys it.
  • Checks titles, descriptions, headings, canonical links, repeated or conflicting tags, structured data and trust cues, each with a printed weight
  • Writes the exact change as a proposal. It has no way to edit your site.
  • Lists what it did not check: page speed, content built by JavaScript, backlinks and rankings
Your website (read-only)robots.txt and sitemap

Limits. Nobody controls rankings, so none are promised. It reads only your own sites and identifies itself honestly.

GEO and AEO / AI search

Found in AI answers

AI search (GEO and AEO) quotes pages that are clear about who you are and what you do. This makes yours easy to quote.

  • Drafts structured data (JSON-LD) from your own product facts, so machines can read your business facts
  • Writes a visible FAQ and the matching FAQ markup, so the markup never says more than the page does
  • Drafts an llms.txt file: an emerging convention with no confirmed ranking effect
  • Proposes which AI crawlers to allow in robots.txt, with the reason for each, for you to decide
Your websiteStructured data (JSON-LD)AI-crawler rules in robots.txt

Limits. Being cited by an AI answer is not something anyone can promise. This only removes the reasons you would be skipped. No citation monitor is connected.

Ads / spend cap

Ads inside a cap you set

Ad copy, audiences and budget moves, held inside a spend limit that is yours.

  • Drafts ad copy and audiences from your real offer, never from invented proof
  • Watches cost per conversion and tells you what is working
  • Moves budget only inside your daily cap, total cap and maximum change per day
  • Stops and asks you the moment a limit would be crossed
Your ad accounts (a sandbox ad platform stands in until you connect one)Your funnel numbers

Limits. No cap set means a cap of zero, and every spend action comes to you. Small samples never steer spend.

Nobody controls where Google or an AI answer places you, so none of this promises a ranking or a citation. It removes reasons to be skipped and shows you what changed. Not offered: app-store optimization, social-media posting, scraping and LinkedIn automation.

What the routine work costs you today.

Staff, times hours, times an hourly cost. Use your numbers.
System to compare against

This is arithmetic on the numbers you enter. It is not a forecast, not a promise and not a client result. Hours given back only count if someone uses them on something that earns.

Questions

Questions buyers ask first.

Is it fully automatic and error-free?

No, and we will not say it is. The agents do the routine work end to end: they read, reply, follow up, quote and chase without waiting for a tap. They also get things wrong sometimes. So failures are retried, then quarantined and shown to you.

Money and limits stay yours. Anything about money runs inside limits you set, and anything above them comes to you. Refunds, payouts and signatures are not things it can do at all. Every action is written to a log you can check, and there is a kill switch.

What does it cost?

Each system has a fixed setup fee and a fixed monthly fee. Setup is $1,500 to $5,000, and monthly is $500 to $2,500 a month. Prices are in USD for businesses outside India. The India list, in INR and excluding GST, is on the currency toggle above.

How long until it is live?

Between 7 and 21 business days from a signed scope and the deposit, depending on the system. The clock starts when the deposit payment lands. The number is on each card, and the card says what is built today.

Which AI does it use, and where does my data go?

The models are Hermes and DeepSeek. The AI Outreach & Lead Qualification Engine runs only on them, the E-commerce Support Assistant and the Connected Operations Layer list them in their stack, and for any other system a person confirms in writing which models are used. Before text reaches a model, email addresses, phone numbers and amounts are swapped for placeholders by pattern, and names are swapped where they have been supplied; that lowers the risk and is not proof that no personal data remains. We have not settled whether hosted models may see redacted personal data, so we do not claim your data never reaches another AI company. Rules and templates decide what happens. The model only helps with wording. Details.

Does it say it is an AI?

Yes. On the first touch, on chat, email and WhatsApp. It is built into the sending step and cannot be switched off. Voice is not built yet; a voice, when there is one, will say it too. If a reply thread already carries the disclosure, it does not stack a second one.

More questions
Who can it message?

Most systems write only to people who wrote to you or agreed to hear from you. The AI Outreach & Lead Qualification Engine is the one that can write first: only to business contacts from public business data, an introduction or an inbound inquiry, each cleared by a consent check, with AI disclosed on the first touch. Whether cold business email is allowed depends on the country; the market pages say what each regulator says. WhatsApp goes through the official Business API: opt-in, plus an approved template outside the 24-hour window. No cold or broadcast WhatsApp, no cold calls, no bought or scraped lists. A STOP is honored at once.

What happens when someone asks about price, a discount or a refund?

Any message with money language goes to the finance desk, never to general sending. The desk answers from the price list and negotiates only inside the limits you set. Above those limits it sends a holding reply and asks you.

Refunds are never decided by the agent. It takes the details, states the policy, and passes it to you.

Why does Connected Operations ask for my approval when the others do not?

Because it works inside your own accounts, across your mailbox, calendar, CRM and documents. For that product, every irreversible action (sending, paying, publishing, deleting) stays human-approved by design. The other systems act on their own, inside the checkpoint and the limits described above.

Do you promise a 100% anything?

One thing only, and it is a count, not a quality claim: 100% of your ICP-fit leads that have a consent record get contacted. The weekly coverage report shows the real number so you can check it. It says nothing about how many reply, book or buy. We do not promise rankings, citations in AI answers, or work without mistakes.

Can I see results from other clients?

Not yet. We will not publish a testimonial, a logo or a result until a real client has earned it and agreed to be named. Until then this page shows what the system can do and how it works, and the live run above is a sandbox on simulated data.

What is llms.txt, and will it get me into AI answers?

It is an emerging convention: a plain-text file that tells AI tools what your site is about. There is no confirmed ranking effect. We publish one because it is cheap and harmless, not because it is proven to help.

Start with the free audit.

Send one web address. In about five business days you get a short written report. If a system fits, a fixed-price scope follows.

Get the free auditSee it run

Free audit, no accounts, we change nothing.

Or ask a person: aimarketing@gonzo.co.in +91 7390990227 WhatsApp

Find your way