How we work together

Five ways this usually starts.

Most work begins with a paid discovery and mapping phase, because a price for a build you have not scoped yet is a guess. Below is what each engagement covers, what it does not, and what we need from you to run it.

01Assessment

Workflow Audit

We walk one existing workflow step by step and return a written list of failure points and repairs, ordered by risk.

Duration: Days, not weeks
Starting point: Smallest commitment

The written output is yours whether or not you continue with us. It is meant to be readable by another engineer.

What it is

Something is already running. It mostly works, and nobody can say how often it does not. We trace the path the work really takes, mark every step that can fail without telling anyone, and write down what each repair would involve.

What you get

A step by step map of the workflow as it runs today
Every failure point we find, ordered by risk
A proposed repair for each one, with rough effort
A short written summary you can forward internally

What we need from you

Read access to the workflow and the tools it touches
One session with the person who handles the exceptions
Any incident history you already keep

Not included

Implementing the repairs
Rebuilding or migrating the workflow
Monitoring after the audit

02First phase of most builds

Discovery and Mapping

The paid first phase of most builds. We map the real process, agree the scope, and write down what success means before anything is built.

Duration: One to two weeks, depending on how many systems are involved
Starting point: Quoted before it starts

If the mapping shows that automation is the wrong answer, we say so and you keep the map.

What it is

Most projects that go wrong went wrong at the scoping stage. Discovery is deliberately paid and deliberately separate, so that finding out in week one that the real problem is somewhere else costs you a small amount instead of a large one.

What you get

A diagram of the current process, including the manual steps
A written scope: what gets automated, what stays human
Success criteria you can check against later
A build plan with sequence and dependencies
An effort range for the build itself

What we need from you

Access to the people who actually run the process
A walkthrough of the tools involved
Someone who can approve the scope when it is ready

Not included

Build work of any kind
A fixed build price before the scope is agreed
A commitment from you to continue

03Build

Focused Automation Build

One clearly scoped workflow or internal tool, built with error handling, alerting, and a runbook from the start.

Duration: Weeks, not months
Starting point: Quoted after discovery

Source and workflow definitions are handed over at the end. Nothing is held back to keep you attached to us.

What it is

A single workflow, not a platform. Error handling is written at the same time as the happy path, alerts go to a named owner rather than a shared inbox, and the runbook is part of the delivery instead of a follow up nobody schedules.

What you get

The working system, in your accounts, under your credentials
Retries with backoff for transient failures
A hold path for permanent failures, with their context kept
Alerting routed to a named owner
A runbook: purpose, dependencies, known failures, recovery
A recorded handoff session

What we need from you

An agreed scope, usually from discovery
Credentials issued under least privilege
A named owner on your side
Test data and somewhere safe to test it

Not included

Work outside the agreed scope, which is quoted separately
Monitoring after handoff, which is Managed Care
Licences and third party subscriptions

04Build

API Product Implementation

Keys, quotas, plans, billing, a dashboard, and documentation. The unglamorous half of shipping an API.

Duration: Weeks to a few months, depending on plan and billing complexity
Starting point: Quoted after discovery

Every claim in this row is checkable on the live product rather than in a slide.

See how API Indonesia was built

What it is

You have data or endpoints and need the product around them. We have shipped this entire stack once already, for API Indonesia: x-api-key authentication, credit quotas, free and paid plans, a self serve dashboard, per endpoint documentation, deployed on Cloudflare. It is live, so you can inspect the result before hiring us to repeat it.

What you get

API key issuance, rotation, and revocation
Credit or request quotas, with the counting rule written down
Plan tiers and rate limits
A billing flow and plan upgrades
A self serve dashboard for keys, usage, and remaining quota
Per endpoint documentation

What we need from you

The endpoints or the data source behind them
A decision on plans and limits, or a session to decide them
A payment provider account and any tax requirements

Not included

Sourcing or licensing the underlying data
Marketing or growing the API
Payment provider fees and infrastructure costs

05Monthly, and optional

Managed Care

Monitoring, credential upkeep, incident response, small improvements, and a scheduled review. Optional, and we would rather say that plainly.

Duration: Monthly, cancellable
Starting point: Monthly retainer

In scope and out of scope are written into the agreement, so an incident is never the moment we work out who pays for what.

What it is

Systems do not stay working on their own. Credentials expire, providers change their responses, and volumes grow past the limits a workflow was designed for. Managed care keeps someone responsible for that. If you have an internal team who will own it, take the runbook and run it yourself. That is a normal outcome, not a failed sale.

What you get

Monitoring on the systems we built, with failure alerts
Credential and token expiry tracking, with renewal reminders
Incident response inside an agreed window
Small improvements within an agreed monthly allowance
A scheduled review of coverage, dependencies, and recovery notes

What we need from you

Continued access to monitor the system
A named contact for incidents
Notice of business changes that affect the workflow

Not included

New workflows or new systems, which are quoted as builds
Systems we did not build, until we have audited them
Third party outages, which we report and work around but do not control
Round the clock on call cover

Budget

We do not publish a price list for this work.

A number on a page would only be honest if every engagement were the same shape. They are not. Cost depends on how many systems are involved, how much of the process is undocumented, and how much of the work is handling exceptions rather than the normal case.

Workflow Audit

Smallest commitment

The least you can spend to find out whether there is a real problem and how bad it is.

Discovery and builds

Quoted after discovery

A build price only means something once the scope exists. Discovery is quoted first, on its own.

Managed Care

Monthly retainer

A fixed monthly figure set against an agreed scope, and cancellable if you no longer need it.

Where we can publish prices, we do

API Indonesia is a product, so its pricing is public and fixed: a free tier of 1,000 requests a month at 10 requests per second, Pro at Rp 49.000 a month for 50,000 credits at 20 requests per second, and Business at Rp 149.000 a month for 250,000 credits at 60 requests per second. Studio engagements are quoted instead, because no two of them are the same size.

Check the pricing on apiindonesia.id

How the number gets set

Pricing is discussed on the fit check, before anyone writes a proposal. You describe what is breaking, we say which engagement fits and roughly what it costs, and if the range does not work for you then nobody has spent anything. Written quotes follow the conversation, never precede it.

Book a Fit Check

Always included

The same four things, in every engagement.

These are not upgrades and they are not line items. They are the reason the work is worth buying, so they are in the scope of an audit and a build alike.

Documentation

A runbook written during the work, covering purpose, dependencies, known failures, and recovery.

Your credentials

Every credential is created in your accounts, under least privilege, and stays yours throughout.

A handoff session

A recorded walkthrough with the people who will own the system, before we call the work done.

A clean exit

Access we hold is listed at handoff and revoked when the engagement ends. Nothing quietly stays open.

Not sure which one

Describe what is breaking and we will tell you which of these fits.

Sometimes the answer is none of them, and we say so. If you want to start, book a fit check and bring one workflow you are tired of babysitting.