Puente OS vs Supabase: Comparison and When to Use Each

A backend is the part of software your team doesn't see: where the data is stored, where it is decided who can sign in and what they can see, where files live and where server code runs. A backend as a service delivers those pieces ready for a development team to build on. Puente OS is the platform that builds the systems that run your business: apps, workflows, databases and agents that read from your systems, cross-reference the information none of them holds in full, decide on your rules and execute the action inside them.

I'm Cris Ugarte, CEO and co-founder of Puente OS. We work with retail, eCommerce and logistics companies in Chile, Peru and Mexico. Supabase comes up in our conversations in two ways: an IT manager who uses it in a product of their own, or someone who got to know it because it sits under an app they built with AI. Either way, the question points to the same piece: where the process's data is going to live, and what it takes around that data for the process to work.

Supabase competes head-on with one of Puente OS's four pieces, the databases, and it is excellent infrastructure. Here I compare the two databases, with public sources for everything I say about Supabase, go through what gets built around the data to run a process, and explain when each one makes sense.

What is Supabase and what does it solve?

Supabase is a development platform built on Postgres, the open source database engine. Its home page opens with the headline "Build in a weekend. Scale to millions." and defines it as "the Postgres development platform: an open source backend for building web and mobile applications." Its GitHub repository is published under the Apache 2.0 license.

What it offers, according to its website and its documentation:

A complete Postgres database. Each project is a dedicated Postgres database, with a table editor, a SQL editor and extensions like pgvector, PostGIS and pg_cron. Its home page describes it as "fully portable," and it generates REST and GraphQL APIs from the schema.
Authentication. Supabase Auth handles user sign-up and sign-in with passwords, magic links, one-time codes, 20+ social login providers and SSO.
Storage. Storage stores, organizes, transforms and serves large files, S3-compatible and with a global CDN.
Realtime. Realtime listens to database changes, syncs user state across clients and broadcasts messages over channels.
Functions and scheduled work. Edge Functions run TypeScript code on the server, distributed globally. Supabase Cron schedules jobs that run SQL, database functions or HTTP requests, and Queues offers durable message queues on Postgres.
Vectors and AI. With pgvector, AI & Vectors stores, indexes and queries embeddings alongside operational data, for semantic, keyword and hybrid search, with integrations like OpenAI, Hugging Face, LangChain and LlamaIndex.
Access rules inside the database. Row Level Security uses Postgres policies to define which rows each user can read or change. Its documentation asks you to enable it on every exposed table and to write a policy for each operation.
Environments and backups. With branching, every change is tested in an isolated copy of the project, which can be connected to GitHub, before it reaches production. Backups are daily and automatic, with up to 30 days of retention, and point-in-time recovery (PITR) restores the database with a granularity of seconds.
Data integrations. Wrappers query external data as if it were Postgres tables: there are wrappers for Stripe, Shopify, HubSpot, Notion, Airtable, Firebase, BigQuery, Snowflake and Slack, among others, plus a generic one for any REST API with an OpenAPI specification. Its partner catalog adds authentication, analytics, AI and development tools.
Tools for AI agents. Its MCP server connects assistants like Claude, ChatGPT, Cursor or VS Code to a project to list tables, run SQL, apply migrations and deploy functions, and its AI tools add reusable instructions for agents. At Supabase Select 2026, on October 2, it introduced an MCP server for the apps built on Supabase, where every call runs with the user's permissions.
Where it runs. In its cloud, with regions in the Americas, Europe and Asia Pacific; in South America, the region is São Paulo. It can also be self-hosted with Docker, and then server operations, updates, backups and availability are the responsibility of whoever hosts it.
Security and certifications. Its security page states SOC 2 Type 2, HIPAA compliance and ISO 27001, with AES-256 encryption at rest and TLS in transit.

Supabase also sits under several tools that build apps with AI, like Lovable and Bolt, which appear in its partner catalog; we look at Lovable in Puente OS vs Lovable.

Standard Postgres, open source, portable data, a large community (its repository has over 110,000 stars on GitHub) and a development team that goes from an idea to an app with sign-in and a database in an afternoon: that is Supabase's job, and it does it very well.

What is Puente OS and what does it solve?

Puente OS is the platform that builds the systems that run your business. Your ERP records what happened. Your BI shows you what happened. Puente makes it happen. In your stack it is the system of action: it reads from your systems, cross-references the information none of them holds in full, decides on your rules and executes the action inside them.

What it builds comes from four pieces, which live on the same platform:

Apps where your team reviews, decides on and approves the exceptions.
Workflows with explicit, auditable rules that start on their own when data changes or at a set time.
Databases that bring together context from several systems: the order, its receipt, its shipment and its settlement in a single record.
Agents for the steps that take judgment, like an ambiguous complaint or a discrepancy that needs investigating.

Each process is organized as a project that groups its apps, tables, workflows and agents, with its own task board. With Puente OS you describe in natural language what your operation needs and get software running: apps, workflows, databases and agents. Puente's Growth Ops team designs each process with you, builds it and maintains it, and each one reaches production in 2 to 4 weeks. Puente acts on your systems, and where the operation lives in spreadsheets, it also keeps the record.

The discipline behind it is autonomous operations design: designing your operation so the repeatable work runs on its own, on your rules, and your team decides what takes judgment. For a closer look at the layer, see what a system of action is.

What does it take around the database to run a process?

A database settles where the data lives and who can touch it. For a process to work, five things get built around it: your business's data model, the connections to your systems, the rules that decide, the screens your team works in and someone who keeps it running when the operation changes. With Supabase, your development team builds those pieces on its infrastructure and chooses where to host each one. In Puente OS, all five live on the same platform as the database.

Take the daily match between a marketplace's sales and the tax documents of a distributor in Chile:

Your business's data model. Someone decides which tables exist and how they relate: the order, its lines, the receipt or invoice that goes with it, the shipment and the marketplace settlement. That design comes from how your company operates, and it defines which questions you'll be able to answer later.
The connections to your systems. Orders come in from Mercado Libre, documents from your invoicing system or Bsale, payments from the bank statement. Each one has its own API, its own credentials and its own format changes, and someone writes and maintains those connections, both ways.
The rules that decide. Most discrepancies explain themselves: a commission, shipping charged separately, a credit note for a return. Those rules resolve the bulk of the cases at any hour, and whatever doesn't match is set aside as an exception.
The screens your team works in. Finance needs an inbox with each exception and its context in view: the order, the document and the payment. There it decides, approves an adjustment or prepares the missing document, with a person's confirmation before it is issued.
Someone who keeps it running. The marketplace changes its API, the business opens a new channel, a rule gets adjusted. Someone sees the change, applies it and leaves the process running.

When those layers live in different tools and with different people, coordinating them eats up your team's hours: that is the coordination tax. An autonomous operation meets three conditions: work starts on its own when data changes, your systems talk to each other with no person in the middle, and the system executes the clear cases and brings you what takes judgment. All three hold up best when the data, the rules, the connections and the screens share a platform. Puente OS delivers all three together, on a single platform.

How do Puente OS and Supabase compare, dimension by dimension?

The first seven dimensions compare Supabase's database head-on with Puente OS's database module. The rest cover what lives around the data.

What the database is. Supabase: a complete, dedicated Postgres database per project, with SQL, extensions and REST and GraphQL APIs generated from the schema; data that is portable to any other Postgres. Puente OS: each process's databases, which run on Postgres, with tables that live on the same platform as its apps, workflows and agents, and that any of them can read and write.
How a table gets created. Supabase: your development team creates it with SQL, in the table editor or with versioned migrations, and can ask an AI assistant connected through its MCP server to do it. Puente OS: you describe what your process needs to store, the Growth Ops team proposes the table structure and you approve it before it is created. A structure change that deletes or converts data asks for your confirmation.
Columns and validation. Supabase: every Postgres type, with relationships between tables, constraints and extensions like PostGIS or pgvector. Puente OS: typed columns for text, number, date, yes or no, and a choice from a closed list, with required fields; the platform validates every row before saving it.
How the data gets queried. Supabase: full SQL, generated APIs and Realtime to listen to changes. Puente OS: queries with filters, grouping and aggregations (sum, count, average, minimum and maximum), the top N in each group and relative dates like today or the start of the month. They are designed for AI to build from a person's question, and the apps, workflows and agents all use them.
Permissions. Supabase: Row Level Security, with row-level policies your team writes for each table and each operation, plus each Postgres role's grants. Puente OS: each company operates in its own space, with its data isolated from other customers' data, and each app accesses only the tables assigned to it, with combinable read, write and delete permissions; the rule is to give it the minimum its work needs: a reporting dashboard gets read only.
Sensitive data. Supabase: AES-256 encryption at rest for all data and TLS in transit, with SOC 2 Type 2, HIPAA and ISO 27001 according to its security page. Puente OS: columns encrypted at rest for sensitive data, defined when the table is created; apps see the encrypted value, and only a workflow from the same work team can read the real value when it needs to.
Loading data. Supabase: imports from CSV, from a Postgres dump or from another Postgres database, and Wrappers to query external sources as tables. Puente OS: loads of up to 10,000 rows in a single operation, which goes in whole or not at all, and workflows that bring the data in from your systems at the right time.
What lives next to the data. Supabase: user authentication, file storage, Edge Functions, Cron, queues and Realtime; your team builds the process screens in the frontend of its choice and hosts them where it prefers. Puente OS: web apps tailored to each process, hosted on Puente with their own link, and workflows and agents on the same platform, grouped in the process's project and working on the same tables.
Integrations. Supabase: Wrappers to query Stripe, Shopify, HubSpot, Notion, Airtable, Firebase, BigQuery, Snowflake or Slack as tables, a generic one for APIs with an OpenAPI specification, and a partner catalog of authentication, analytics, AI and development tools. In its full documentation and its partner catalog, as of October 13, 2026, there are no integrations for Mercado Libre, Bsale, Defontana, the SII (Chile's tax authority) or Falabella; those services connect through their API with code your team writes, for example in an Edge Function. Puente OS: native connectors for Mercado Libre (orders and shipping details), Bsale (stock and documents), Multivende, WooCommerce, WhatsApp, Gmail and Google Sheets, among others. The Growth Ops team connects the rest, like the SII, your invoicing system or your ERP, through its API, over HTTP or Python, or by reading the ERP's database.
Work that runs on its own. Supabase: Edge Functions, Cron to schedule SQL, functions or HTTP requests, durable queues and Realtime to react to database changes; your team writes the logic of each job. Puente OS: workflows that start from a webhook or a schedule, with explicit rules, iterators and per-step failure policies, and agents inside them for the steps that call for judgment, all reading and writing the process's tables.
AI. Supabase: pgvector for semantic and hybrid search over your data, an MCP server so your development team's AI assistants can work on the project, and instructions for agents. Its documentation recommends connecting them to production only when the task requires it and keeping manual approval of each action. Puente OS: the AI is in how the whole system gets built, in natural language, and in the agents that take the judgment steps inside each workflow.
Versions and governance. Supabase: branching with preview environments per pull request, versioned migrations, daily backups and point-in-time recovery. Puente OS: every workflow save creates a new version, and rolling back means activating the previous one; in apps, every publish is saved as a version, with earlier ones kept as history. Every execution is recorded with the step where it stopped and, per step, the error, the traceback and the logs, and irreversible actions, like issuing a tax document, are designed in two steps: a draft and a person's confirmation.
Where it runs, who operates it and who responds. Supabase: in its cloud, in the region you choose, with São Paulo as the option in South America, or on your own infrastructure with Docker; Supabase operates the platform, and the app you build on top is maintained by your development team. Puente OS: Puente manages the platform where everything runs, and the Growth Ops team maintains each process as your operation changes, with a new channel, a new rule or a report that changes format, and responds when something fails.

What does it look like in a real operation?

It looks like data from several systems brought together in tables that work for a process, with the rules, screens and agents on the same platform and a team that maintains it. Two Puente OS customers:

Example 1: a multichannel electronics seller in Mexico, where Puente keeps the operation's record

An electronics seller in Mexico runs without an ERP, so the record of its operation lives in Puente's databases, connected to seven of its sales channels. On top of those tables, every order follows its path on rules: the address is validated, the carrier is chosen by zone, cost and insurance, the label is generated and the tracking number is written to the order. A report consolidates 213,362 Mercado Libre orders via API, and a competitor price capture has collected 1,209,812 records and saves the team 33 hours a month.

Of the 38,628 orders processed between November 2025 and August 2026, 97.9% got their shipping label with no human intervention: 37,815 labels. Generating the labels for a cutoff went from 5 to 7 hours to 15 minutes, and monthly volume grew 40% with the same operations team. From August 1 to 12, 12 of 2,051 orders had an incident, 0.59%, with a median resolution time of 10 minutes. In all, that is 20 apps, 75 processes, 7 channels and 19 APIs, built in ten months by one person.

Example 2: a pet retailer in Chile, where cross-referenced data comes back as actions

A pet retailer sells on four channels. Its OMS (order management system) on Puente processed 56,597 orders between January and August 2026, and every shipment is measured against the delivery promise: 33,192 shipments measured and 85.0% on-time delivery among those not canceled, calculated against the courier's API. A workflow cross-references sales with their costs every day and calculates the margin of every line: as of August 2026 it had covered 247,626 lines, and 3,830 of them, 1.5%, showed a negative margin that nobody saw before. What gets decided on that data goes back to the store: 8,362 price changes executed in its online store from Puente.

It runs 10 apps in production for six departments, with 167 versions deployed, built and maintained by one person. The time from requesting a new process to having it in production dropped from 105 days, on the first project, to 7.

Why choose Puente OS?

Because you build the application end to end on a single platform, and you get it working. Supabase is excellent infrastructure for a development team to build its product on Postgres, with open source code and portable data; in Puente OS, the database, the apps, the workflows, the agents and the integrations with your ERP, your eCommerce, your marketplaces and your invoicing live together, with a single team in charge: Growth Ops designs it around your processes, builds it and maintains it. Six concrete reasons:

From the database to the running process, on a single platform. Each process's tables, apps, workflows and agents live in the same project and work on the same data, with one provider and one team in charge. At the pet retailer in Example 2, that adds up to 10 apps in production for six departments, with 167 versions deployed.
The tables are born inside the process. Growth Ops models the data on how your company works, with typed columns validated before saving and read, write and delete permissions for each app. Where there is no ERP, Puente's databases keep the record: at the electronics seller in Example 1, which runs without an ERP, that adds up to 38,628 orders processed end to end.
The data arrives connected to your systems, both ways. Puente has native connectors for Mercado Libre, Bsale, Multivende and WooCommerce, and the team connects your invoicing system, your courier or your ERP through its API, over HTTP or Python, or by reading the ERP's database. At the pet retailer, 8,362 price changes were executed in its online store from Puente.
The rules and agents work on those same tables. Workflows with explicit rules handle the bulk of the work, agents take the steps that call for judgment and every execution is recorded step by step. At the electronics seller, 97.9% of orders go out with their label with no human intervention.
A team builds it, maintains it and responds. Growth Ops puts each process in production in 2 to 4 weeks, adjusts it when your operation changes and responds when something fails. At the electronics seller, from August 1 to 12, 2026, 0.59% of orders had an incident, with a median resolution time of 10 minutes.
Each new process builds on what's already there. The next process uses the tables, connections and rules already running on the platform, and it is also described in natural language. At the pet retailer, the time from requesting a process to having it in production dropped from 105 days to 7.

That way your operation runs on your rules, and your team spends its time deciding what takes judgment.

When does each one make sense?

Supabase fits when you have a development team building a product of its own that wants solid backend infrastructure on Postgres. Puente OS fits when you want your business process working end to end on a single platform, connected to your systems, with a team that builds it with you and maintains it.

Supabase, if your development team is building a web or mobile app for your customers, a SaaS product or a platform of your own, and needs a database, user sign-in, files, realtime and server-side functions from day one. Also if you want full Postgres with SQL, relationships and extensions, portable data, the option to host it on your own infrastructure, vector search for AI features or certifications like SOC 2, HIPAA and ISO 27001. Its open source code, its community, the speed it gives a development team and the portability of Postgres are real strengths.
Puente OS, if you want the process left running with its tables, its screens, its rules and its agents on the same platform, connected to Mercado Libre, Bsale, your invoicing system or your ERP, with a team that designs it with you, maintains it and responds when something fails. Also if your team would like to spend its time deciding the exceptions, and you want the next process to be added on the same platform.

How do you start?

With a process that hurts, with volume and manual handoffs: generating shipping labels, reconciling a marketplace or matching documents. With Puente OS, each process reaches production in 2 to 4 weeks.

Pick the process. With volume, manual handoffs and a visible cost when it fails.
Take the baseline. Cases per month, hours of manual work per week and the share of cases that come out right the first time.
Take inventory of where the data lives. The databases, spreadsheets, apps and scripts that touch that process, with what they store, who maintains them and which systems they feed. That list defines what gets connected and what gets modeled again.
Connect the sources. Your ERP, your store, your marketplaces, your invoicing system, your courier and your bank statements.
Write down the rules your team applies from memory. For example: "if the receipt for a Mercado Libre order matches the sale, the order is reconciled; if the document is missing or the amount differs, it goes to finance with the order and the payment in view."
Decide where each exception goes. The system resolves it, or it reaches a named person with full context in an App.
Measure from the first week and add the next process on the same platform.

That is autonomous operations design applied process by process: the repeatable work runs on its own, on your rules, and your team decides what takes judgment. To choose which steps go to a rule and which to an agent, this guide on how to implement AI in your business can help, and if you're evaluating tools that build apps on top of a backend like Supabase, see Puente OS vs Lovable and Puente OS vs v0.

We believe every company in Latin America can be operationally autonomous, and that starts with having every process running end to end, with its rules and a team that keeps it running.

Key takeaways

Supabase is an open source development platform built on Postgres: a dedicated database, authentication, storage, realtime, Edge Functions, Cron, queues and vector search, for teams building web and mobile apps.
Supabase adds Row Level Security, branching, backups with point-in-time recovery, an MCP server for AI assistants, the option to self-host, and SOC 2 Type 2, HIPAA and ISO 27001.
Supabase competes head-on with Puente OS's databases. Puente's run on Postgres and have typed columns validated before saving, read, write and delete permissions per app, queries with aggregations and encrypted columns for sensitive data.
Around the data, a process needs your business's model, the connections to your systems, the rules that decide, the screens your team works in and someone who keeps it running.
Puente OS delivers all of that end to end on a single platform: tables, apps, workflows, agents and integrations, designed around your processes and built and maintained by the Growth Ops team. At an electronics seller in Mexico, 97.9% of orders go out with their label with no human intervention.
Supabase fits when a development team is building a product of its own. Puente OS fits when you want the process running end to end, in 2 to 4 weeks per process.

Want to see what your data looks like working inside a process? Let's talk about your processes, and we'll show you Puente OS running with cases like the ones in this article. Book a conversation with our team.

Frequently asked questions

What is Supabase and what is it used for?

Supabase is an open source development platform built on Postgres that gives a development team the backend of an app: a database, user authentication, file storage, realtime, server-side functions and vector search. It's used to build web and mobile apps, SaaS products and AI features on your own data, and it sits under several tools that build apps with AI. It runs in its cloud or self-hosted on your own infrastructure.

What is a backend, explained for a manager?

It's the part of software you don't see. When someone signs in to an app, the backend checks who they are, shows them only what they're allowed to see, stores what they enter and runs the server's calculations. Supabase delivers that part ready for a development team to build on. To run a business process, around the data you need your operation's model, the connections to your systems, the rules, the screens for your team and someone who maintains it. Puente OS delivers all of that together, on a single platform and in production.

What are Puente OS's databases like?

They run on Postgres, they are one of Puente OS's four pieces, together with apps, workflows and agents, and they live on the same platform. Each company operates in its own space, with its data isolated from other customers' data. Growth Ops models them on your operation: tables with columns for text, number, date, yes or no, and a choice from a closed list, validated before saving; queries with filters and aggregations; loads of up to 10,000 rows in a single operation, and columns encrypted at rest for sensitive data. Each app accesses only the tables assigned to it, with read, write and delete permissions. They bring together context from several systems in a single record, and where the operation lives in spreadsheets, they keep the record the action needs; your existing systems, like the ERP, remain the source of truth for every transaction.

Does Supabase integrate with Mercado Libre, Bsale or the SII?

In its full documentation and its partner catalog, as of October 13, 2026, there are no integrations for Mercado Libre, Bsale, Defontana, the SII or Falabella. Its Wrappers cover services like Stripe, Shopify, HubSpot, Notion or Airtable, and a generic one queries REST APIs with an OpenAPI specification; the rest connects with code your team writes and maintains, for example in an Edge Function. Puente OS has native connectors for Mercado Libre (orders and shipping details), Bsale (stock and documents) and Multivende, and the Growth Ops team connects the rest via API.

What happens to what we already have on Supabase?

It goes into the inventory. The Growth Ops team reviews with you the databases, apps and scripts that touch each process, what they store and who maintains them, and decides which ones stay where they are, like your product's backend, and which processes get designed within Puente OS. Supabase provides a standard Postgres connection for external tools, and Puente has a PostgreSQL connector that runs parameterized queries, so a workflow can read that data when the process needs it. Each process reaches production in 2 to 4 weeks.

External sources cited

Everything this post says about Supabase comes from its website, its documentation, its blog and its GitHub repository, reviewed on October 13, 2026. Puente OS customer figures are production measurements as of August 2026; label generation time, the hours saved on price capture and the request dates for each process come from follow-up with the customer.

Supabase: product

Supabase, home page: headline "Build in a weekend. Scale to millions."; "Supabase is the Postgres development platform: an open source backend for building web and mobile applications"; a dedicated, portable Postgres database with REST and GraphQL APIs, authentication with 20+ social login providers, S3-compatible storage, Edge Functions, Realtime and vectors with pgvector; open source tools that can be self-hosted; MCP server.
Supabase on GitHub: "The Postgres development platform"; Apache 2.0 license; about 111,000 stars as of October 13, 2026.
Supabase documentation: database, authentication, storage, realtime and Edge Functions; AI & Vectors, Cron, Queues, Data REST API and GraphQL; libraries for JavaScript, Flutter, Python, C#, Swift and Kotlin; CLI and Management API.
Database overview (Supabase documentation): every project gets a full Postgres database; table editor, SQL editor and extensions like pgvector, PostGIS and pg_cron.
Auth (Supabase documentation): passwords, magic links, one-time codes, social login and SSO.
Storage, Realtime and Edge Functions (Supabase documentation): large files with access policies; listening to database changes, shared state and channels; globally distributed server-side functions.
Cron and Queues (Supabase documentation): scheduled jobs on pg_cron that run SQL, database functions or HTTP requests; durable message queues on pgmq.
AI & Vectors (Supabase documentation): pgvector for embeddings, semantic, keyword and hybrid search; integrations with OpenAI, Hugging Face, LangChain, LlamaIndex and Amazon Bedrock.
Row Level Security (Supabase documentation): authorization rules that run inside the database; enable it on every table in an exposed schema and write policies per operation.

Supabase: platform, integrations and AI

Branching (Supabase documentation): isolated environments per branch, preview branches tied to GitHub pull requests, persistent branches and migrations on merge.
Database Backups (Supabase documentation): automatic daily backups with up to 30 days of retention; point-in-time recovery with a granularity of seconds.
Regions (Supabase documentation): regions in the Americas, Europe and Asia Pacific; in South America, South America (São Paulo), sa-east-1.
Self-Hosting (Supabase documentation): Supabase hosted with Docker on your own infrastructure; operations, updates, backups and availability are the responsibility of whoever hosts it.
Connect to your database (Supabase documentation): direct connection, a pooler for external tools and the Data API.
Foreign Data Wrappers (Supabase documentation): querying external sources as Postgres tables. The full documentation (supabase.com/llms-full.txt, reviewed on October 13, 2026) includes Wrappers for Stripe, Shopify, HubSpot, Notion, Airtable, Firebase, BigQuery, Snowflake, Slack and APIs with an OpenAPI specification, among others, and does not mention Mercado Libre, Mercado Pago, Bsale, Defontana, Nubox, Falabella or the SII.
Supabase partners: the partner integration catalog, including Lovable and Bolt, reviewed on October 13, 2026; it does not include Mercado Libre, Bsale, Defontana, Nubox, Falabella or the SII.
Supabase MCP server (Supabase documentation): AI assistants connected to a project to list tables, run SQL, apply migrations and deploy functions; the recommendation to connect production only when the task requires it and to keep manual approval of each action.
AI tools (Supabase documentation): MCP server, reusable instructions for agents and plugins.
Supabase Select 2026 Recap (Supabase blog, October 2, 2026): schema and configuration in code, native local development in alpha, an MCP server for apps built on Supabase with each user's permissions, health alerts and multi-node high availability in private alpha.
Security (Supabase): SOC 2 Type 2, HIPAA compliance, ISO 27001, AES-256 encryption at rest and TLS in transit.

Puente OS vs Supabase: From the Database to the Running Process, on a Single Platform