the system understands the operation, sources and constraints
Applied intelligence · Juiz de Fora, Brazil
AI that understands the work before automating it.
Synap designs artificial intelligence and automation systems for complex operations. We connect context, evidence and action so teams can move forward with greater clarity, traceability and control.
outputs remain connected to their origin and can be reviewed
people keep authority over consequential decisions
What we do
We turn operational complexity into usable systems.
Our work starts with the decision and the workflow—not with a fashionable model. Technology follows the problem, the evidence and the level of risk.
Applied AI
Models, retrieval and agents designed around real tasks and measurable outcomes.
Intelligent automation
Workflows that connect systems, rules and people without hiding exceptions.
Operational knowledge
Structured context, source lineage and memory with explicit access and retention rules.
Responsible delivery
Human review, clear boundaries, observability and safe exits built into the product.
How we work
Understand. Design. Prove. Scale.
Understand the operation
We map decisions, sources, exceptions, responsibilities and the cost of getting them wrong.
Build with evidence
We create a controlled path from input to result, with review points proportional to risk.
Improve from real use
We observe what happened, measure quality and evolve the system without losing traceability.
03 · Autonomous agents
Autonomy to coordinate. Control to build trust.
We work with autonomous-agent architectures that can understand goals, consult sources, use authorized tools and coordinate specialized steps — from analysis through execution.
Autonomy proportional to risk.
An agent is never given unrestricted freedom. Permissions, memory, evidence and validation checkpoints define what may proceed independently and what requires a human decision.
Orchestration
Specialists operating as one system.
A goal is decomposed, assigned to specialized agents and brought back together after tools are consulted and evidence is validated.
Operational trust
Memory, boundaries and a trace for every action.
Persistent context, explicit permissions and checkpoints follow the path from intent to result.
How an agent moves forward
01
Understand
Reads the goal, context and constraints.
02
Plan
Breaks down work and coordinates specialists.
03
Act
Uses authorized sources and tools.
04
Validate
Returns evidence and review checkpoints.
Critical or external actions remain under the authority of the people accountable for the operation.
Our direction
Built in Brazil to raise the standard of applied AI.
Synap is a Brazilian startup with the ambition to help lead the country’s progress in artificial intelligence and automation. We pursue that ambition through useful products, disciplined engineering and trust earned in real operations.
05 · Co-founders
Executive leadership · Juiz de Fora, Brazil
Synap is built by people who understand complexity from within.
Dalton Rodrigues and Camilla Dutra lead Synap as co-CEOs and co-founders. Their experience in complex legal operations meets artificial intelligence development, product thinking and automation.
01Co-CEO & Co-founder
Dalton Rodrigues
Attorney, Civil Procedure specialist, CEO and co-founder of Synap Inteligência Artificial. His work brings together legal expertise, artificial intelligence, Machine Learning and technology product design for real-world problems. At Synap, he connects domain knowledge, technology and product thinking to make work more organised, traceable and ready to move forward.
Education and credentials
- Specialist in Civil Procedure
- IBM AI Developer Professional Certificate
- AI, Machine Learning and AI & Law · Lund University, University of London, Vanderbilt University, DeepLearning.AI, Google and IBM
Continuous AI educationDalton’s artificial intelligence certifications22 credentials · 2024
Law, AI and engineering brought together in a continuous technical foundation.
Dalton's learning path spans AI and Law, language models, generative AI, Machine Learning, data and software development. The catalogue preserves the title, issuer, date and identifier supplied for each credential.
View credentials on LinkedInAI and Law
02- AI & LawLund UniversityIssued Dec 2024 · Credential M809OFYDGCRC
- Generative AI for Legal Services PrimerVanderbilt UniversityIssued Oct 2024 · Credential O9793OPSTRQQ
Generative AI, LLMs and prompting
11- Generative AI: Elevate your Software Development CareerIBMIssued Nov 2024 · Credential ICBQ8N904KXY
- Building Generative AI-Powered Applications with PythonIBMIssued Nov 2024 · Credential RS289MYZAT4A
- Introduction to Large Language ModelsGoogleIssued Oct 2024 · Credential GYYL9RHLVCJA
- Introduction to Responsible AIGoogleIssued Oct 2024 · Credential GQVATQFSKFVA
- Introduction to Generative AIGoogleIssued Oct 2024 · Credential S4V85ZR9CXMC
- Introduction to Artificial Intelligence (AI)IBMIssued Oct 2024 · Credential RQNV46G9QM69
- Prompt Engineering for ChatGPTVanderbilt UniversityIssued Oct 2024 · Credential HX5R5KR6KT0R
- Generative AI: Prompt Engineering BasicsIBMIssued Oct 2024 · Credential QPIBIU93LBNI
- Generative AI PrimerVanderbilt UniversityIssued Oct 2024 · Credential QL570T6PSPFI
- Generative AI: Introduction and ApplicationsIBMIssued Oct 2024 · Credential OW6TXL7IBZTB
- Inteligencia Artificial (IA): Interacciones y promptsUniversidad de PalermoIssued Oct 2024 · Credential F8GTFUJ5SHQF
Machine Learning and data
04- Supervised Machine Learning: Regression and ClassificationDeepLearning.AI · Coursera · Stanford CPD · UVMIssued Nov 2024 · Credential YMXK77WHZFH3
- Machine Learning with PythonIBMIssued Oct 2024 · Credential Y7N9K64MPRBR
- Fundamentos: dados, dados, em todos os lugaresGoogleIssued Oct 2024 · Credential IJJJFNFTDGO7
- Machine Learning for AllUniversity of LondonIssued Oct 2024 · Credential OV91PCKX1AK9
Development and engineering
05- IBM AI Developer Professional CertificateIBMIssued Nov 2024 · Credential Q7M55DNO9X0A
- Developing AI Applications with Python and FlaskIBMIssued Oct 2024 · Credential I6C8JYOGDKME
- Python for Data Science, AI & DevelopmentIBMIssued Oct 2024 · Credential XVKFMI5ZBFJZ
- Introduction to HTML, CSS, & JavaScriptIBMIssued Oct 2024 · Credential X7JKCUFOMDVQ
- Introduction to Software EngineeringIBMIssued Oct 2024 · Credential CJSD7SBBT7R8
02Co-CEO & Co-founder
Camilla Dutra
Attorney, with a Law degree from the Federal University of Juiz de Fora (UFJF), CEO and co-founder of Synap Inteligência Artificial. She specialises in Civil Procedure, Labour, Public and Social Security Law, with extensive experience in high-volume litigation, high-risk legal portfolios and tax enforcement. At Synap, she connects technology with operational reality to turn legal challenges into clear, useful products.
Expertise and experience
- Law degree · Federal University of Juiz de Fora (UFJF)
- High-volume litigation
- High-risk legal portfolios
- Tax enforcement proceedings
Synap Journal
Research, product thinking and the signals that matter.
Original analysis lives separately from the daily AI and Machine Learning Radar. Sources, dates and authorship remain visible by design.
Bring us an operation that deserves to work better.
We will start with the context, identify where intelligence creates leverage and define a responsible next step.