Lucas Resende Tavares
Solutions Architect & Sales Engineer at RavenDB
Nearly two decades writing .NET software and modeling data, from Delphi cursors to vector search over documents. Today I help engineering teams, technical decision-makers and business owners trade complexity for predictability, performance and productivity, worldwide, with a focus on the Americas, Europe and Oceania.

Start here
Message me on WhatsApp
Pick a topic and WhatsApp opens with the message ready. You just add your case. That way I answer with context on the first reply, without trading five questions first. No form, no sign-up.
I want to evaluate RavenDB for my project
Tell me the scenario and I will say, with technical honesty, whether RavenDB is the right call.
Open WhatsAppCommercialI need a quote or licensing
I align the technical side with you and bridge you to the RavenDB commercial team.
Open WhatsAppSupportI already use RavenDB and need technical help
Slow query, heavy index, unstable cluster: let us look at it together.
Open WhatsAppProof of conceptI want a POC with my own data
A real slice of your use case, benchmarked against the path you are on today.
Open WhatsAppMigrationMigrating from another database to RavenDB
From relational or another NoSQL, with the current system staying up.
Open WhatsAppCommunityInvite me to a talk or meetup
Conference, community, university or podcast, on-site or remote.
Open WhatsAppTrainingTraining for my team
Document-oriented modeling, indexes, search and AI over your data.
Open WhatsAppArchitectureDiscuss architecture and modeling with RavenDB
How to design the document model, the indexes and the architecture around RavenDB.
Open WhatsAppEnglish or Portuguese, in your time zone
A technical conversation with no middle layer, with an answer within the same business day.
Direct line to engineering
When a question goes beyond me, it reaches the RavenDB engineering team in Poland.
Brazilian companies in production
RavenDB already runs in production in Brazil, from fintech to industry, and several of those projects went through me.
Prefer e-mail? Write to lucas.tavares@ravendb.net
What I have delivered, in numbers
4–8 h → 40 s
Critical procedures
Logic rebuilt with table functions and cursors removed, without changing the data model or adding a single new index.
1.8 min → 600 ms
Automotive DaaS
Technical recommendation engine remodeled on RavenDB with Full-Text Search. Authentication included in that response time.
30 → 45 days
Negotiated and met
The client wanted it in 30 days. I asked for 45, rebuilt the entire architecture as documents and delivered in 45. By the first Friday it already compiled.
About me
From SQL to documents: what I learned along the way
I was born in Belo Horizonte and have lived in Curitiba since 2015. I've been programming since 2007, when I wrote my father's company website in Notepad, and professionally since 2009, immersed in Delphi, Firebird and SQL Server. I spent my first years learning to understand data in order to optimize the business — not the other way around.
In 2018 I was hired to audit a system the client wanted live in 30 days. The constraint was non-negotiable: RavenDB, a database I had never heard of. I asked for 45 days. In the first week I rebuilt the whole structure as documents, and I delivered within the 45. That was the end of the migrations soap opera and the endless joins. Back then there was no AI to help me: it was Stack Overflow, documentation, Google and elbow grease!
Since then I have designed software with RavenDB whenever the choice was mine. I only used other databases when the client had a clear infrastructure constraint. I applied it to performance-critical projects, published open-source libraries for the ecosystem and took the topic to stages like TDC and FAE. In 2024, at Developer Week, I met Oren Eini and David Baruc in person — and in August 2025 I officially joined the RavenDB Ltd team.
I do not advocate for RavenDB because I work there. I work there because I chose RavenDB as my tool of trust for almost a decade, solving real production problems with it.
What I believe
Everything I write follows the same thread: almost every expensive technology bill is a complexity bill. Five tools to keep one thing running, a service that fails and takes three others down, a cloud invoice nobody can explain. Simplifying takes courage. And it is usually the most profitable decision a technical team can make.
That is why I argue for document data, text and vector search in the same layer, and AI applied to the whole of your data, not just to one corner of it. When an older system has to keep running, CDC is one of the routes; it is not the point, just one way in. It is what I have seen work in production, at scale and under pressure. My mission is to help companies make technology decisions that make them more profitable, with faster, simpler and smarter applications.
Off the keyboard: I climb and hike, listen to rock and metal, and run a 1984 Belina turned beer truck, Choppellina. I also write science fiction: Lambda X will be my first book.
My favorite song is called The Raven Child, by Avantasia. Yes, there's a raven in the name. A coincidence, but I adopted it.
Open on SpotifyStory
Why RavenDB? The full story
A college itch, a system nobody wanted to tackle in 30 days and an international phone call. Here's how I got to RavenDB and became its bridge to Brazil.
How I help with RavenDB
Sales engineering is engineering first. I understand your team's problem and show, with running code, whether RavenDB solves it. When it does not, I say so.
01
Evaluation and proof of concept
You describe the scenario; I model a real slice of your data in RavenDB, benchmark it against your current path and tell you, with technical honesty, whether it's worth it.
02
Modeling and production performance
Document-oriented modeling, indexes, projections and cluster review: from 'it's slow' to 'why it's slow', with a plan your team can execute.
03
Legacy migration, CDC and ETL
From relational or another NoSQL to RavenDB, while the current system stays up. CDC and ETL when legacy can't stop.
04
Search, vectors and AI over your data
Full-text search, vector search and AI in the same data layer, applied to your whole dataset, not just one end of it.
Through RavenDB, I work remotely from Brazil for clients anywhere, in English and Portuguese, and on site when the project calls for it (as in San Francisco, San Jose and São Paulo). I've worked with teams in the US, Europe and Israel, so time zones aren't an issue.
Stack and specialties
Career
From technical support to RavenDB’s global team.
2025 — Present
RavenDB Ltd
Sales Engineer · Solutions Architect
Technical and commercial support for companies evaluating or scaling RavenDB: modeling, proofs of concept, performance, legacy migration and AI features. Working with clients in Brazil and abroad.
2018 — 2025
LeadSoft Soluções Web
Founder, CEO and .NET Backend Developer
Founded and led a software house focused on custom software and high performance, official RavenDB partner in Brazil between 2024 and 2025. Ran teams, contracts and architecture for clients such as Castrol Brazil, ECX Pay, Inforlube and Ebbi Tecnologia.
2015 — 2018
Mastertech Sistemas
Mid-level Developer · Delphi and .NET C#
Evolution of enterprise management products, integrations and maintenance of relational databases in production.
2013 — 2015
Banco Bonsucesso · BHS
Junior Developer · Delphi and .NET C#
SQL Server query and script optimization, merge review and version control, plus .NET development with NHibernate.
2011 — 2013
Leadmec Industry & Trade
Junior Delphi Developer · DBA
Requirements gathering, systems analysis and building the internal ERP in Delphi 2010, including Firebird database administration.
2009 — 2011
Arquimedes Informática
Delphi Developer · intern and trainee
The first years: support, system maintenance and the discovery that performance is a design decision.
Education
B.Sc. in Information Systems — Faculdade Cotemig (2010–2015) · IT Technician (vocational high school) — Colégio Cotemig (2007–2009)
Languages
Portuguese (native) · English (fluent, technical and conversational)
Based in
Curitiba, Brazil. Remote and on-site, in Brazil and abroad
Focus: the Americas, Europe and Oceania
Previous chapter
LeadSoft Soluções Web
Between 2018 and 2025 I founded and led LeadSoft, a software house in Curitiba and an official RavenDB partner in Brazil between 2024 and 2025. That's where theory became practice: contracts, teams, deadlines and the responsibility of sustaining what I architected myself. In 2025 I wound down the software house and took what I learned to the RavenDB team.
Clients such as Castrol Brazil, ECX Pay, Inforlube and Ebbi Tecnologia, plus open source libraries published to make patterns like Repository and Guids easier inside the Raven ecosystem.
Por isso, podemos afirmar que a Leadsoft® figura entre os Parceiros "Top Five" do Grupo Inforlube.
That is why we can say that Leadsoft® is among the "Top Five" Partners of the Inforlube Group.
Remo Lucioli, Inforlube Group (translated from Portuguese)
Software development (.NET) and technology and product consulting remain available on request.
Talk about development or consultingPast client? Leave a review on GoogleOpen source
My open source libraries
MIT-licensed open source .NET packages, that I run in production and other teams do too. They came out of problems I had: generic repositories, identity on RavenDB, and integrations with Brazilian public APIs plus AWS, Azure and Google.
208.855
NuGet downloads
17
Published packages
MIT
Open source license
.NET core
Common.Library
A Swiss-army knife of extensions, middlewares, enumerators and utilities to speed up any .NET application.
.NET core
Common.GlobalDomain
Reusable global-domain entities and classes, the DDD foundation I carry into every project.
.NET core
Common.DataAccess
Data-access layer with multi-database support, built to swap providers without rewriting the repository.
Brazilian APIs
Adapter.BrasilAPI
Testable .NET adapter for BrasilAPI public services, built around IHttpClientFactory.
Brazilian APIs
Adapter.ViaCep
Brazilian postal-code lookup through ViaCEP with a light, injectable, test-ready layer.
Brazilian APIs
Adapter.IBGE
Integration with the Brazilian IBGE public APIs: cities, states and geographic data.
Identity & e-mail
Adapter.Email
Transactional e-mail services configured through Dependency Injection.
Identity & e-mail
Adapter.RavenDBIdentity
ASP.NET Identity running on RavenDB: authentication as documents, no relational tables.
Cloud
Adapter.Aws.SecretsManager
AWS Secrets Manager through Dependency Injection: secrets out of appsettings.
Cloud
Adapter.Google.ReCaptcha
Google reCAPTCHA validated on the backend, with configurable policy and threshold.
Cloud
Adapter.Google.Workspace
E-mail delivery and Google Workspace integrations from service accounts.
Stages and expos
Events I have been to
Technology conferences and expos in Brazil and the United States. I get on stage and talk to people who run data every day.
Want to invite me? Bio, photos and materials are on the press page.

Developer Week
San Francisco, USA
Met David Baruc, Oren Eini and the RavenDB team in person: the turning point of my career.

Castrol · Turbo Mode ON
Mogi das Cruzes, Brazil
Pre-launch talk for the Castrol Trocador app (Android and iOS), for all Castrol distributors in Brazil.

TDC São Paulo
São Paulo, Brazil
Talk and workshop: NoSQL, a small step in modeling, a giant leap in efficiency.

FAE · Entrepreneur Week
Curitiba, Brazil
Talk "O efeito borboleta no empreender" (the butterfly effect in entrepreneurship), for students: from a question to a professor to RavenDB, and why to pay attention to the subtleties of college.

Microsoft Ignite
San Francisco, USA
First expo already on the RavenDB team, focused on .NET architecture and data in the cloud.

TechEx North America
San Jose, USA
Live demos at the RavenDB booth for Silicon Valley data and AI teams.
Video
On video
Talks, workshops and demos on modeling, performance and RavenDB.
NoSQL: a small step in modeling, a giant leap in efficiency
Talk and live demo at TDC São Paulo 2024: how small modeling changes produce massive performance gains. In Portuguese.
Watch on YouTube →The Power of RavenDB — NoSQL, Vector Database, Generative AI and AI Agents
How a document database comes to power generative AI and agents: vector search, embeddings and hybrid queries in the same data layer.
Watch on YouTube →5 reasons to use RavenDB as THE database of your product
Five technical reasons to put RavenDB at the center of your product rather than at the edge of the architecture.
Watch on YouTube →RavenDB Playground: test performance without installing anything
A tour of the public playground: run queries, indexes and Full-Text Search straight from the browser, no infrastructure required. In Portuguese.
Watch on YouTube →What I write: data, architecture and the complexity bill
Articles I published on LinkedIn, almost always with code and numbers.

Nobody told you that you can have AI without migrating your database
I simulated a legacy system nobody is allowed to shut down (a SQL Server with years of history) and layered AI, smart search and a conversational agent on top of it, live, through RavenDB CDC. Without touching a single line of what was already there, with an open demo anyone can try. In Portuguese.
Read on LinkedIn
RavenDB + AI: real-time unified segmentation
Documents, text search and vector search in one query. Why keeping database, search engine and vector store as three separate products gets expensive, and what changes when they all live in the same data layer. In Portuguese.
Read on LinkedIn
Lessons from processing 4% of a country’s GDP
What I learned building and running software at a scale where downtime has an immediate price, and why agility there is not a slogan. In Portuguese.
Read on LinkedIn
When simplifying takes courage: the hidden price of complexity
Every extra tool in the architecture carries a cost that never shows up on the invoice: a team held hostage, fragile integrations, nobody able to explain what was modernized. In Portuguese.
Read on LinkedIn
Stop using class for everything in C#: a simple guide
When class, struct and record actually change allocation, performance and the shape of your application, explained without academic detours. In Portuguese.
Read on LinkedIn
Why I talk so much about RavenDB
The story behind the choice: what made me leave relational out of conviction rather than fashion, and why I started publicly arguing for a simpler data stack. In Portuguese.
Read on LinkedInCommunity
I take RavenDB to stages and communities
TDC São Paulo, GDG Maringá, Developer Week in San Francisco, Microsoft Ignite. A talk, a workshop, a community meetup, a university class or a podcast episode: if people want to understand data modeling, vector search or AI over data, I show up.
Let’s talk
If you deal with data, architecture or technology budgets, we probably have something to talk about. Pick a channel.
Lucas R. Tavares
Sales Engineer · RavenDB
Full career history, articles and whatever I am publishing right now.
Connect on LinkedInFast answers, straight to me. Best for scheduling a technical conversation.
+55 41 99182-1569
Official RavenDB
Keep up with RavenDB