Software engineering studio · Brazil → Worldwide

We build thesoftware youroperation runs on.

Custom web platforms, APIs, workflow automation, and data & AI products — designed, engineered, and shipped end to end, from discovery to production.

Founded by João Paulo Santos
Live platforms and products
04
Project write-ups
14
Bilingual delivery
EN · PT
Overlap with US & EU hours
UTC−3
  • TypeScript
  • Python
  • Django
  • SvelteKit
  • React
  • Astro
  • Tailwind CSS
  • PostgreSQL
  • Neon
  • Supabase
  • Cloudflare Workers
  • Docker Swarm
  • Three.js
  • deck.gl
  • Mapbox
  • XGBoost
  • Stripe
  • HTMX
  • Sentry
  • Playwright

Web platforms, APIs, Automation, Data, Geospatial, Applied AI, Cloud, DevOps

01Services

Engineering for the systems that matter to your business.

One team from architecture to deployment. We take on the hard parts — data models, integrations, reliability — so your product ships and keeps working.

01

Custom web platforms

SaaS products, portals, and internal tools built for real operational load.

  • SaaS & multi-tenant apps
  • Institutional portals
  • Internal tools & back-offices
Learn more: Custom web platforms
02

APIs & integrations

Clean contracts between your systems, with guarded failure modes.

  • REST APIs & BFFs
  • Payment & business-system integrations
  • Webhooks & event pipelines
Learn more: APIs & integrations
03

Workflow automation

Replace spreadsheets and manual hand-offs with pipelines that run themselves.

  • Process automation
  • Scheduled jobs & ETL
  • Reporting & reconciliation workflows
Learn more: Workflow automation
04

Data, geospatial & ML

Turn weather, map, and operational data into decisions.

  • Geospatial apps (Mapbox, deck.gl)
  • Forecasting & risk models
  • Dashboards & KPIs
Learn more: Data, geospatial & ML
05

Applied AI

LLM features that improve a real task — not a generic chatbot.

  • LLM features & agents
  • Source-grounded retrieval
  • Usage metering & cost controls
Learn more: Applied AI
06

Cloud, DevOps & reliability

Edge-first deploys, containers, CI, and observability from day one.

  • Cloudflare Workers & Pages
  • Docker Swarm & Portainer
  • CI/CD with automated tests
Learn more: Cloud, DevOps & reliability

02Selected work

Software built for real operations.

Platforms for agribusiness, education, innovation ecosystems, and operations teams — each with a documented problem, decision, and outcome.

Explore all projects
  1. 2026

    Moto Track

    A live SvelteKit SaaS web application for motorcycle operations, covering fuel, maintenance, documents, costs, professional shifts, and profitability with Neon and Cloudflare.

    Live SaaS

    Product status

    Capabilities

    • Fuel, maintenance, tires, documents, reminders, and expenses
    • Professional shifts, work costs, and profitability reports
    • SvelteKit
    • TypeScript
    • Neon Postgres
    • Cloudflare
    • Stripe
    Moto Track — preview
  2. 2026

    AgroHub UniRV

    A live Django web portal for UniRV's startup ecosystem, events, services, and public-access channels, built as a fast, maintainable institutional home.

    Startups · events · services

    Portal scope

    Capabilities

    • Public directory of incubated startups
    • Events calendar for mentorships, workshops, and pitch days
    • Python
    • Django
    • HTMX
    • Tailwind CSS
    • Alpine.js
    AgroHub UniRV — preview
  3. 2026

    Inova Rio Verde

    A live geospatial web application that turns Rio Verde's hubs, incubators, and programs into an explorable citywide innovation trail.

    Citywide innovation trail

    Product surface

    Capabilities

    • Interactive geospatial 'Innovation Trail' mapping the local ecosystem
    • Edge-optimized Postgres access via Cloudflare Hyperdrive
    • Next.js
    • Tailwind CSS
    • Mapbox
    • MapTiler
    • deck.gl
    Inova Rio Verde — preview
  4. 2026In development

    Lorebound

    A reader-first interactive-fiction web platform that pairs curated story worlds with controlled AI turns, persistent memory, and transparent usage.

    Story packs

    Reader model

    Capabilities

    • Curated worlds set the tone before an AI turn extends it
    • Reader choices and memory inform the next chapter
    • TypeScript
    • React
    • Neon Postgres
    • Drizzle
    • Cloudflare Workers
    Lorebound — preview
  5. 2026Private source

    ClimAgro

    ClimAgro UniRV, coordinated by Prof. Dr. Gilmar Oliveira Santos, brings together weather data and climate analysis to support agricultural planning in Rio Verde.

    Daily + hourly

    Weather resolution

    Capabilities

    • FAO-56 evapotranspiration and water-balance calculations with historical warm-up
    • Paired INMET producer and Django consumer with contract-drift checks
    • Python
    • Django
    • JavaScript
    • PostgreSQL
    • INMET
    ClimAgro — preview
  6. 2026

    Throughline

    A privacy-first React PWA planner that keeps goals, tasks, notes, boards, and timelines usable offline, with optional encrypted sync across devices.

    Local data remains usable

    Offline behavior

    Capabilities

    • Goals roll up task progress instead of sitting beside disconnected to-dos
    • Notes cross-link to tasks and goals
    • React
    • TypeScript
    • IndexedDB
    • PWA
    • End-to-End Encryption
    Throughline — preview

03Process

A delivery process you can see every week.

No black boxes. Scope is explicit, decisions are written down, and working software lands on a staging URL from the first sprint.

  1. Step 011–2 weeks

    Discovery & scope

    We map the operation, the users, and the constraints, then turn them into a scoped plan with priorities, risks, and a clear first milestone.

    Outputs

    • Technical scope
    • Risk map
    • Milestone plan
  2. Step 021–2 weeks

    Architecture & design

    Data model, integrations, and interface flows are designed before the build, so there are no expensive surprises halfway through.

    Outputs

    • Data model
    • UX flows
    • Decision log
  3. Step 03Weekly cadence

    Build in sprints

    Weekly increments on a live staging environment. You review working software every week, not status slides.

    Outputs

    • Staging URL
    • Weekly demos
    • Automated tests
  4. Step 04Ongoing

    Launch & evolve

    Production release with monitoring and documentation, then either a clean hand-off to your team or continuous evolution with us.

    Outputs

    • Production release
    • Runbooks
    • Support plan

04Under the hood

Engineering standards, not afterthoughts.

The quality bar that ships with every project — the same one this website is built on.

// edge/api/orders.ts — typed, validated, observableimport { z } from 'zod';import { withTracing, problem } from '@/lib/platform'; const Order = z.object({  customerId: z.string().uuid(),  items: z.array(LineItem).min(1),}); export default withTracing(async (request, env) => {  const body = Order.safeParse(await request.json());  if (!body.success) return problem(422, body.error);   const order = await env.DB.orders.create(body.data);  await env.QUEUE.send({ type: 'order.created', id: order.id });   return Response.json(order, { status: 201 });});
typecheckunite2ea11ydeploy → edge
  • Type-safe end to end

    Strict TypeScript and validated boundaries between client, API, and database.

  • Performance budgets

    Static and edge-first delivery with Core Web Vitals tracked in production.

  • Accessible by default

    WCAG 2.1 AA as the target: keyboard paths, contrast, and reduced-motion support.

  • Observable in production

    Error tracking, structured logs, and real-user metrics from the first deploy.

  • Tested pipelines

    Unit and end-to-end tests run in CI before anything reaches production.

  • Secure by design

    Least privilege, bot protection, and encryption where user trust depends on it.

Reference architecture

L1Interface

  • Astro
  • SvelteKit
  • React
  • HTMX
  • Tailwind

L2Application

  • Django
  • Workers
  • REST / BFF
  • Queues

L3Data

  • Postgres
  • Neon
  • Supabase
  • IndexedDB

L4Infrastructure

  • Cloudflare
  • Docker Swarm
  • CI/CD
  • Sentry

05Sectors

Domains we already know.

Context shortens projects. These are the domains we have built for — live platforms, research, and prototypes, each with its status stated on the case study.

  • Agribusiness & climate

    Agronomy, farm finance, and climate-risk platforms that turn field and weather data into decisions.

  • Innovation & education

    Portals and ecosystems for universities, incubators, and public innovation programs.

  • Operations & services

    Software for individual riders and field work, plus a customer self-service prototype for service companies.

  • Digital products & AI

    Consumer products with privacy, offline capability, and AI at the core.

06Engagement

Pick the model that fits the stage you are in.

Every engagement starts with a short discovery call and a written proposal.

  • 01

    Scoped project

    Best for: New products, portals, and MVPs

    A defined scope delivered in milestones, with discovery included and a complete hand-off at the end.

    • Discovery included
    • Milestone-based delivery
    • Source code & documentation hand-off
    Discuss this model
  • 02Recommended

    Product partnership

    Best for: Products that keep evolving

    Dedicated monthly capacity to build and evolve your product continuously, with a shared roadmap.

    • Weekly delivery cadence
    • Shared roadmap ownership
    • Priority support
    Discuss this model
  • 03

    Technical consulting

    Best for: Existing systems that need a plan

    Architecture reviews, audits, and recovery of stalled projects, ending in an actionable plan.

    • Code & architecture audit
    • Performance & accessibility review
    • Prioritized remediation plan
    Discuss this model

07FAQ

Questions companies usually ask.

How big is the team?

JPCLOW is founder-led by João Paulo Santos. When a project needs additional expertise, independent specialists can be brought in for a defined scope.

What kinds of companies do you work with?

Startups validating a product, established companies modernizing internal processes, and institutions that need reliable public-facing platforms — in Brazil and abroad.

How does a project start?

With a short discovery call. Then you receive a written proposal with scope, milestones, timeline, and a clearly defined first deliverable.

Can you take over an existing system?

Yes. We start with a technical audit, stabilize the critical paths, and then evolve the product incrementally instead of proposing a rewrite by default.

Which technologies do you use?

We choose per problem. Our defaults are TypeScript, Python and Django, SvelteKit, React, Astro, PostgreSQL, and Cloudflare.

Do you work remotely and in English?

Yes. We are remote-first from Brazil (UTC−3) and work in English and Portuguese, with working-hour overlap for teams in the Americas and Europe.

How do I request a proposal?

Use the contact form or email [email protected] with a short description of the problem, your timeline, and any existing systems involved.

Next step

Have a system to build or a process to fix?

Tell us about the problem. You get a first technical read, a suggested path, and next steps.