Real-time Analytics · Risk Management
BBVA's trading desk needs to respond to institutional client requests for bond, interest-rate swap, and foreign-exchange quotes within milliseconds while balancing competitiveness, profitability, portfolio risk, and regulatory control.
Reported impact
BBVA reports broader coverage of quoted products, faster response times, more competitive pricing, and more cost-effective, personalized solutions for institutional clients while maintaining model robustness and regulatory compliance.
BBVA describes an in-house AI stack for algorithmic trading that blends predictive models and generative AI into daily trading-desk workflows. Client quote requests arrive through Bloomberg and are evaluated against live market conditions, hedge liquidity, dealer participation, and the desk's current risk exposure. BBVA's algorithms combine those real-time signals with historical trading data to calculate prices and support hedging and execution decisions in milliseconds. The bank's ePricer capability adds a real-time personalization layer by producing "On the fly ideas" tailored to each client's needs. For user-facing delivery, BBVA embedded Viriato, an AI assistant, inside its Atalaya platform for fixed-income sales teams and traders. Before release, models are backtested on historical evidence, then assessed with A/B testing in real-world conditions, with Risk and Control involved to maintain robustness and regulatory compliance.
Institutional clients send quote requests through Bloomberg. BBVA's trading models combine historical trading evidence with real-time market signals, hedge liquidity, dealer participation, and current portfolio risk to calculate prices and execution decisions in milliseconds. ePricer generates tailored ideas in real time, and Atalaya with the embedded Viriato assistant delivers AI support to fixed-income sales teams and traders under Risk and Control governance.
"A year ago, the bank took its first step with Viriato, an assistant integrated into the Atalaya platform for fixed-income sales teams and traders."
"AI also makes it possible to offer more cost-effective and personalized solutions, such as ‘On the fly ideas’, an ePricer feature that generates proposals tailored to each client’s needs in real time"
"Algorithms combine historical data with real-time signals to calculate a quote that strikes a balance between competitiveness, profitability and risk control."
"Before being deployed into production, each model is validated against historical evidence through backtesting; its performance is then assessed under real-world conditions using techniques such as A/B testing."
"Clients see direct benefits: broader coverage of quoted products, faster response times and more competitive pricing."
"teams rely on advanced models to analyze vast volumes of data, optimize pricing and streamline execution."
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
The Organization JSON-LD lists streetAddress 'Plaza de San Nicolas 4', addressLocality 'Bilbao', addressCountry 'ES', and numberOfEmployees value 87534.
Related
In-house Sample Apr 16, 2026 UBS Asset Management UBS Asset Management needs to support multi-asset investing decisions across strategy design, tactical asset allocation, portfolio construction, security selection, research, and portfolio implementation while rapidly… Automated trade scheduling · Dendrogram clustering · Execution optimization algorithms In-house Sample Jul 28, 2026 Novartis Novartis R&D needs to turn large volumes of scientific, clinical, operational, and real-world data into faster, higher-quality decisions across drug discovery and development, reducing cycle times, surfacing safety… less than 10 percent of drug candidates historically reach the market AI models · AI-enhanced analytics · Generative AI 
Google Cloud Sample Archived · Apr 13, 2026 Finhike Finhike needed a cloud platform that could keep a lending stack fast and available in unstable connectivity environments, power real-time risk decisions, and let the team launch country-specific services with minimal… BigQuery · Cloud Logging · Cloud Monitoring 