MANOEL SOUZA DA SILVAMANOEL S. SILVA
Cascavel, PR, Brazil · GMT-3 · in operations since 2017

I turn operational problems into systems, data and automation that deliver results.

Before the code
0+
Years inside operations before building the systems they run on.

I did not start by building software. I started by running the processes software had to solve: shipping, FX, reconciliation, distribution, supplier audits.

Today I bring together official trade data from several countries, at scale and with no manual work, and deliver it as a tool the sales team uses to decide.

See the work Live projects Career
Pipeline · automated monthly collection● active
MDICFile + API
USDA FASOfficial API
US CensusOfficial API
EurostatBulk file
GACCCustom collector
UN ComtradeOfficial API
WarehousePostgreSQLnormalizedincrementalaudited
Official records in a single warehouse
0
✓ VERIFIED · DATA PLATFORMMDIC · USDA · CENSUS · EUROSTAT · GACC · COMTRADE
0International government sources, collected automatically every month✓ VERIFIED · DATA PLATFORM
0Official datasets from six origins read the same way✓ VERIFIED · DASHBOARD PROJECT
0Straits monitored live over AIS✓ VERIFIED · FAROL PROJECT
0+Years inside operations, from 2017 to today✓ VERIFIED · RESUME
The order mattersI lived every step before automating it.
01 Operations
02 Business rules
03 Data
04 Systems
05 Audit
06 Decision
The work · real cases

Six cases.
One method.

In color: the step where my work comes in.
Foreign trade company · 2024 to present

Six official sources, one data model, 270,000,000 records.

“Comparing markets cannot start with fixing formats.”

01 · Problem

Each country publishes in its own tariff code, currency and calendar. Chinese customs has no API. Every comparison took weeks: dataset by dataset in Excel, NCM converted to HS one code at a time.

02 · Solution

A PostgreSQL warehouse with MDIC, USDA FAS, US Census, Eurostat, GACC and UN Comtrade, nomenclatures normalized into one model, custom collectors where there is no API and an automated monthly load.

03 · Result

From weeks of manual work to an automated monthly load. The decision on where to sell now comes from a single dataset.

Official sourcesCollectionNormalizationValidationWarehouseDecision
✓ VERIFIED · RESUME
01 / 06 · auto-advancing
Live projects

Four projects.
Live.

Portfolio versions of problems I solved at work, on open or demo data. You can open and use them; they run on a free server, so the first visit can take up to a minute.

Automatic updatesLive

Foreign Trade Dashboard

Monthly trade balances that update themselves.

The trade balances that matter to agribusiness, month by month and updated automatically: exports and imports of meat, fruit and grain across 26 official datasets from six origins, read the same way.

  • Brazil, United States, Europe, China, Colombia and Mexico
  • 9,000,000 rows in Parquet, read by DuckDB
  • Comparisons only between periods with the same months
PythonStreamlitDuckDBPlotly
Open the project
01 / 04 · auto-advancing
In production, with a real client
Smart Fit purchasing platforms

For the chain's units in Brazil, Mexico and other Latin American countries. Where units and personal trainers buy products and uniforms. I coordinated development, with inventory integrated with Texport's ERP, online payment and a direct link to the distribution center I set up.

smartfitstore.com.brSmart Fit StoreThe chain's products and uniforms store, with inventory integrated with Texport's ERP.Visit the sitestorepersonal.com.brStore PersonalUniform purchases by the units' personal trainers.Visit the site
The sources behind the work

Ten public, official datasets.

Collected, normalized and updated with no manual work.

01 / 10File + API
MDIC · Comex Stat
Ministry of Development, Industry, Trade and Services · Brazil

Every Brazilian export and import, by month, product, country and state of origin.

What it shows▸FOB value and net weight▸Destination or origin country▸Exporter state and municipality
Format8-digit NCM · US$ · kg
UpdatesMonthly, with revision of the latest month
Where I use itData platform, Dashboard and Garimpo
02 / 10Official API
US Census · USDA FAS
U.S. Census Bureau and Foreign Agricultural Service · United States

United States exports and imports, with the USDA agricultural cut.

What it shows▸Value and quantity by product▸Partner country and customs district▸FAS agricultural series
Format10-digit HTS · US$
Updatesmonthly
Where I use itData platform and Dashboard
03 / 10Bulk file
Eurostat · Comext
Statistical office of the European Union

Trade between EU countries and with the rest of the world.

What it shows▸Value and weight by product▸Reporter and partner country▸Imports at CIF value
Format8-digit CN · € · kg
Updatesmonthly
Where I use itData platform and Dashboard
04 / 10Custom collector
GACC
General Administration of Customs · China

Imports and exports of China, the largest buyer of Brazilian protein and grain.

What it shows▸Value and quantity by product▸Origin and destination country▸Monthly series by chapter
FormatHS · US$
UpdatesMonthly, collected automatically
Where I use itData platform and Dashboard
05 / 10Official API
UN Comtrade
United Nations Statistics Division

The multilateral base: trade reported by almost every country, in one standard.

What it shows▸Flows between any pair of countries▸Value and quantity▸Comparison with national data
Format6-digit HS · US$
UpdatesMonthly and yearly, as each country reports
Where I use itData platform
06 / 10Bulk file
Federal Revenue · CNPJ
Brazilian Federal Revenue Service

The open registry of every company in the country.

What it shows▸Registration status and opening date▸Main and secondary activities (CNAE)▸Size, share capital and partners
FormatCNPJ · CNAE · IBGE municipality
Updatesmonthly
Where I use itProspecting and registry checks, Garimpo
07 / 10Boundaries API
IBGE · boundaries
Brazilian Institute of Geography and Statistics

Official boundaries of states and municipalities.

What it shows▸Geometry of every state and municipality▸Official municipality code▸Base for the supply map
FormatGeoJSON · 7-digit IBGE code
UpdatesWhen IBGE updates the boundaries
Where I use itProspecting by region, Garimpo
08 / 10Official catalog
MAPA · SIF
Ministry of Agriculture and Livestock · Brazil

Establishments under federal inspection, the ones allowed to ship animal products.

What it shows▸SIF number and company name▸Field of activity▸Municipality and state
FormatSIF number · CNPJ
UpdatesWhen the catalog changes
Where I use itExport readiness score, Garimpo
09 / 10WebSocket
AIS
Signal from the ships themselves, via the AISStream receiver network

Position, heading and speed of ships with AIS transponders, near the coast.

What it shows▸Position every few seconds▸Name, type and dimensions▸Declared destination and arrival
FormatMMSI · IMO · knots · degrees
UpdatesReal time
Where I use itTracking cross-checked with trade data, Farol
10 / 10SPARQL query
Wikidata · Commons
Wikimedia Foundation

The open history of hulls, identified by IMO number.

What it shows▸Build year and shipyard▸Tonnage and operator▸Photo with author and license
Format7-digit IMO · SPARQL
UpdatesOn demand, cached
Where I use itEach ship's profile, Farol
Career · 2017 to present

No layer replaced the one before.

Each phase added a skill to the ones already there. That is why the system ships with the right business rules.

2017201920212023today
Operations: finance, logistics, audit, trade
Scale: SAP, WMS and network distribution
Structure: distribution center and in-house software
Data and systems: warehouse, ERP, audit
2017Phase 01 · OperationsAdvant ConfecçõesFrom checker, picker and packer to Administrative Assistant
Feb 2017 to Aug 2021

I joined in 2017 loading, checking and picking. That same year, being trusted with Smart Fit's Black line became a promotion, and I went through finance, HR, sales and logistics.

■The first pilot exports of my career, orchestrated by me, to Smart Fit units in Peru, Colombia and Ecuador
■Feasibility study (cost, inventory, value transfer and responsibilities) presented to Smart Fit's board, the basis for moving from São Paulo to Paraná
■Full finance cycle: payables and receivables, reconciliation, cash flow and monthly P&L
■Supplier and factory audits, including ABVTEX certification
2021Phase 02 · ScaleLojas Quero-QueroAdministrative Assistant · loss prevention and distribution operations
Aug 2021 to Jul 2022

I learned that systems at scale fail on details, and that errors have to show up before production.

■Distribution center from 57 to 550 stores
■Administration of Körber's HighJump WMS, integrated with SAP ERP
■PALAVRA program: loss and damage prevention, and about 18% lower operating cost
■Distribution center shipping with its own team: routing, deliveries and pickups
2022Phase 03 · StructureTexportSales Assistant · international contracts, distribution and logistics
Aug 2022 to Feb 2024

Back in the group, now at headquarters. I set up an operation from scratch and, for the first time, built the software it runs on.

■Smart Fit's distribution center and the web purchasing platforms for units and personal trainers
■Items across several categories for the Brazilian Paralympic Committee at the Santiago 2023 Parapan American Games, from tender to shipment
■End-to-end supply chain for Puma, Asics, Fila, Mizuno and other brands
2024Phase 04 · Systems and dataForeign trade companyMarket Intelligence Analyst · data platform, trade and systems
Feb 2024 to present

Today I bring together the operation I know, the business rules and official data in a system that proves its own results.

■Data platform with six government sources and 270,000,000 records
■ERP for four companies with finance on a single ledger
■Warehouse audit that found three silent failures
■Intranet used by every employee, replacing WhatsApp and keeping history for audits
Current trading company · 2024 to present

Systems for five layers of the same company.

DataForeign trade collection and warehouse
IntelligenceProspecting and market analysis
OperationsSales, trade, logistics and finance ERP
ManagementInternal tools and integrations
CommunicationIntranet replacing WhatsApp, with history for GPTW and ISO
How I think

Five rules I follow.

Understand the operation before the codeBefore modeling, I learn how the business really works. Most of the time I have already done that job by hand.
Business rules come before the databaseThe data model represents the real rule, not the structure most convenient for the system.
Automate the repetition, expose the exceptionIf information moves by hand from one system to another, I ask why, and the exception becomes an alert.
Silent errors are the worst errorsA system that delivers a wrong number without warning is worse than a slow system.
Prove it against real dataMigrations, automations and KPIs are checked against production data before they count.
COMING SOON
Personal project · in validationClara · AI Secretary + CRM

In international trade, many contacts that arrive on WhatsApp are lost for lack of time to reply and follow up. Clara answers in Portuguese, English and Spanish, by text and voice, captures product, volume, destination, Incoterm and target price, looks up NCM codes, generates a preliminary proforma PDF, books meetings and hands off to the team when needed. Every conversation becomes a lead in a custom CRM.

+Text and voice support in 3 languages

Replies in the contact's language, understands incoming voice notes and replies by voice with the agent's cloned voice.

+Meetings booked in Google Calendar

Checks the team's free slots, books, reschedules and cancels meetings and sends a reminder beforehand.

+CRM with profile, score and next step

A second AI reads each conversation, builds the contact and product profile, scores the deal potential and suggests the next step. Duplicate contacts are merged.

+Automatic follow-ups, reminders and alerts

Follows up on stalled contacts, reminds people of meetings and alerts the team about unanswered customers, with a daily summary every morning.

COMING SOON
Personal project · in progressControle Financeiro

A desktop app for net worth and investments that works offline and keeps everything on your machine.

+Returns by XIRR and TWR

XIRR measures return using the date and amount of every deposit and withdrawal; TWR measures portfolio performance without the distortion of those cash flows.

+OFX and CSV statements with no duplicate entries

Imports bank and broker statements and recognizes what was already recorded, so nothing is entered twice.

+Recurrences suggested, never created on their own

Spots expenses and income that repeat and suggests adding them; nothing is saved until you confirm.

About

From operations to systems.

manoel-pb.webp
Manoel Souza da Silva

I did not start by building software. I started by running the processes software had to solve: shipping, FX, reconciliation, distribution, supplier audits.

Today I bring together official trade data from several countries, at scale and with no manual work, and deliver it as a tool the sales team uses to decide.

Each phase added a skill to the ones already there. That is why the system ships with the right business rules.

0+Years inside operations, from 2017 to today
0Official datasets from six origins read the same way
0Live projects, open to use
Talk to me
Frequently asked questions

Before we talk.

What kind of problems do you solve?

Operational problems that become systems, data and automation: an official international trade data platform, ERP, prospecting, warehouse auditing and internal tools.

Which technologies do you work with?

In the projects on this page: Python, PostgreSQL, DuckDB, Streamlit, Plotly, Altair, Node, CesiumJS and MapLibre.

Are the projects live?

Yes, all 4 projects are live and you can open and use them. They run on a free server, so the first visit can take up to a minute; this page wakes the servers up while you read.

Are the numbers on this page real?

Yes. Every number on this page has a source, the CV or the project itself, marked as checked next to it.

Where are you and how do I reach you?

In Cascavel, PR, Brazil (GMT-3). By email, phone, LinkedIn or GitHub, in the contact section below.

Got a problem worth solving?
Let's talk.

manoelsouzadsilva@hotmail.com
+55 45 99921-8991
LinkedIn GitHub
Manoel Souza da Silva · Cascavel, PR, BrazilEvery number on this page has a source.