Firebase development

Build scalable, real-time mobile and web applicationswith Firebase and Google's infrastructure

Firebase can give a small team authentication, a real-time database and hosting without a server to run. The catch is in the details: access rules live in your hands, bills follow your read patterns, and leaving later takes work. We plan for those from the first day.

What Firebase work covers

Most of the value, and most of the risk, is in the data model and the rules.

  • Firestore data modelling

    Structuring collections around the screens that read them, because Firestore queries are limited and every document read is counted.

  • Security Rules

    Client apps talk to the database directly, so the rules are your access control. We write them with tests, using the local emulator.

  • Cloud Functions

    Server-side work such as payments, notifications and third-party calls, for anything a client app can't be trusted to do.

  • Cost and performance reviews

    Finding listeners and queries that read far more documents than the screen needs, which is a common reason for a bill rising unexpectedly.

Firebase details to plan around

From Google's Firebase documentation, as of October 2026.

  • Budgets don't stop spending

    Google is explicit that budget alerts don't cap your usage or charges. They only notify you. Cloud Functions needs the pay-as-you-go Blaze plan, so a spending safeguard has to be part of the design.

  • Write limits are real

    Google recommends the 500/50/5 rule: at most 500 operations per second on a new collection, growing 50% every 5 minutes. Sequentially indexed fields such as timestamps limit a collection to 500 writes per second.

  • Some products have ended

    Firebase Dynamic Links was scheduled to shut down on 25 August 2025. An app still relying on those links needs another deep-linking route. Check every Firebase product you rely on.

When Firebase is right

It is very good at some things, and hard to escape when it stops fitting.

  • A good fit

    Mobile-first products, chat and live updates, and early products with a small team and no one to run servers. Authentication and notifications alone can save real time.

  • A poor fit

    Heavy reporting, complex joins and financial records across many documents. Firestore is not designed for ad hoc queries, so reporting usually needs an export to something else.

  • SQL Connect is an alternative

    Firebase Data Connect, now called SQL Connect, provides a managed PostgreSQL database through Cloud SQL, with GraphQL and generated SDKs. It suits relational data within Firebase.

  • Plan the exit in advance

    Firestore data and rules don't move directly to another database, and app code often calls Firebase everywhere. Keeping those calls behind a small layer makes leaving much cheaper.

How a Firebase project runs

The rules and the bill are designed with the data, not added on afterwards.

  1. List screens and their reads

    For each screen, we estimate document reads and writes, so you can see a realistic bill before you commit to a structure.

  2. Write rules alongside the model

    Rules and data structure are designed together, and rule tests run in the emulator as part of CI.

  3. Separate the environments

    Development, staging and production are different Firebase projects, so that a test can't touch real users or real data.

  4. Launch with alerts and App Check

    Budget alerts, a spending safeguard and App Check to limit abuse are all in place before the app is public.

How Firebase work is scoped

An audit of rules and costs is a sensible first step for an existing app.

A Security Rules and cost review is short and well defined, so it can be quoted up front. A full app build is scoped from the screen list. Hourly help runs from $20/h in USD, invoiced in USD, GBP, EUR or INR.

We are an engineering team in India with 3–5 hours of overlap with the UK and Europe; US calls by appointment. A mobile build and its back end can be tested together each day. If you expect relational data or heavy reporting, compare this with a conventional back end on Node.js or Laravel before you decide.

Why this page lists no Firebase apps

Most of our work is delivered under NDA, so many projects cannot be named.

Accordingly, we explain what we would look at in your project, and leave client names out. Sharing your Security Rules file with us is a practical first step, and one that tells us a lot.

Questions about Firebase

Often asked once an app has gone live and the bill arrives.

Why is our Firebase bill rising?
Usually because of too many document reads: listeners on large collections, screens re-fetching data or missing pagination. Google's usage reports show where it goes, and restructuring those screens generally fixes it.
Are default Firebase rules safe?
Rules written during development often leave data too open, and can stay that way by accident. Any public app needs its rules reviewed and tested. This is the first thing we check.
Can we move off Firebase later?
Yes, but it takes real work. You export the data, rebuild the rules as a back end and replace the client calls. Authentication users can also be exported. It is much easier if planned for at the outset.
Firestore or Realtime Database?
Firestore is the usual choice for richer queries and structured data. Realtime Database can suit very frequent small updates to simple data. We compare them against your actual read and write pattern.

Worried about your rules or your bill?

Tell us what the app does and which Firebase products you use. We will tell you what we would review first.

We reply within one business day (IST).