guibvieira

- Led the design and deployment of Echo, a decentralized Telegram data aggregation protocol using smart contracts to coordinate 40–50 ingester nodes. - Developed orchestrator contracts for node registration and IPFS data hash submission, enabling modular dapps to process decentralized social data. Founded -- TrendMoon, a data analytics layer built on Echo that processes 20k+ Telegram groups in real time. Led pipeline development for ingestion, spam filtering, segmentation, summarization, and API delivery. - Built AI-driven summarization and segmentation agents to extract coin-specific community insights, narrative shifts, catalysts, unlocks, and sentiment. Designed and executed social-data-informed trading strategies leveraging mindshare, sentiment, and attention signals. - Led staking optimizations and performance modeling for Taraxa’s native DPoS liquid staking system. - Supported business development and ecosystem growth by integrating oracles, advising on integrations, and onboarding new projects.

Blockchain Energy Research Group Member

Delivered ML-based PoC for BFSI clients including frontend development and backend recommendation systems.

• Using Scala and Spark to aggregate Big Data, performing ETL to transform AML, transactional data into document-based data it in order to load into Elasticsearch. • Aggregation, processing and migration of data on large bank datasets depending on clients needs to fix their data inconsistencies in order to help users extract useful insights in financial crime when using the platform. • Administrate performance tests to the company platform using Gatling Framework to stress test Elasticsearch clusters up to 40 nodes using Kubernetes and helm for deployment. • Optimise elastic search cluster configurations for performance, implement snapshots and restoration mechanisms. Achieved this with the help of monitoring tools like Grafana, Kibana and GCP tools. • Development of new score implementations to enhance or better detect AML and Tax Evasion behaviours from network data and document data using aggregated world wide datasets with transactional data from the largest Danish Bank. • Using Jupiter Notebook and Python for Data and Graph analytics from the score implementation models used on the networks. Developed analytical toolkit library to extract and visualise insights about customers/business behaviour as well as model tuning for better AML detection and fulfilment of compliance requirements.

• Inventory management and process optimisation for better ground service support. • Had the opportunity to suggest a new automated way of inventory management in order to solve an existing problem in inventory management, developed an implementation plan and presented to senior stakeholders.