Vinicius Fernandes

Co-founded Ambern, a small AI/automation engineering studio positioned as "2 engineers for startups" for international clients. Focused on WhatsApp AI agents, billing/collections automation, and RAG-based systems. Running this in parallel with client work — open to new full-time opportunities.

Built data-driven backend systems and AI-assisted workflows for B2B operations.
• Developed lead-generation and data-processing workflows using Python, PostgreSQL and public business datasets.
• Built an LLM Agents–powered CRM automation using LangChain and LangGraph to orchestrate multi-step WhatsApp Business API workflows.
• Developed backend integrations connecting APIs, database state and busi

Designed and implemented data cross-referencing and enrichment pipelines for financial data via Open Finance, and developed AI/LLM pipelines including RAG for risk analysis and scoring.

Designed and implemented data cross-referencing and enrichment pipelines for financial data (Pix, accounts, credit) via Open Finance, on a platform serving 3,000+ active clients, mapping 74M+ companies across 30+ data sources with 80% validated accuracy.
Developed AI/LLM pipelines — including Retrieval-Augmented Generation (RAG) — for risk analysis, scoring, and insight generation from consented t

Architected batch and near real-time data pipelines supporting financial intelligence and regulatory compliance systems.
Key Contributions:
• Optimized high-volume SQL queries and Python processing jobs, reducing overall pipeline execution times by ~85%.
• Implemented automated data quality validation frameworks across core database tables, improving system reliability for audit readiness.
• Partn

Led the evolution of checkout and payment systems, improving conversion rate by 18%, and built purchase-behavior and LTV data pipelines with AI solutions for fraud detection and product recommendation.

Led the evolution of checkout and payment systems (Pix + card), improving conversion rate by 18%.
Built purchase-behavior and LTV data pipelines, cross-referencing payment data with user engagement.
Implemented AI solutions for fraud detection and product recommendation (lightweight models + feature store).
Stack: Node.js/TypeScript · React · PostgreSQL · Kafka · Python · DBT · AI applied to fraud

Developed and maintained high-volume data pipelines for Pix transaction processing and reconciliation, implementing key cross-referencing with Receita Federal databases and reducing reconciliation time by 40%.

Developed and maintained high-volume data pipelines for Pix transaction processing and reconciliation (millions of events/day).
Implemented Pix key cross-referencing with Receita Federal databases and internal antifraud systems.
Optimized queries and data modeling in PostgreSQL and Cassandra, reducing reconciliation time by 40%.
Contributed to the evolution of event-driven architecture (Kafka) for

Focused on database systems, data engineering and backend development, with practical application in AI-powered software systems.

Coursework and independent engineering projects cover relational data modeling, SQL, database architecture, Python, data pipelines, REST APIs and software development.

Alongside the degree, I build production-oriented projects involving PostgreSQL, FastAPI, Docker, LL