BridgeWise, a global leader in AI for wealth, has unveiled major upgrades to Bridget™, its AI chat platform designed specifically for financial markets.
As financial institutions in Thailand increasingly adopt artificial intelligence, Bank of Thailand have placed policy guidelines on AI risk management such as data quality, model explainability, and cybersecurity threats that make it harder for financial institutions to navigate AI implementation.
To assist with this, BridgeWise has designed Bridget™ with a strong focus on responsible and enterprise-grade AI. Built on Amazon Web Services (AWS), the platform incorporates multi-layered accuracy validation, explainable reasoning workflows, and secure data handling to address critical challenges in financial AI, including accuracy, explainability, regulatory readiness, and the mitigation of AI hallucinations.
A recent webinar hosted by Elad Nachmias, Chief Technology Officer at BridgeWise, together with Yaniv Gueta, Principal Solutions Architect – Financial Services at Amazon Web Services (AWS), offered an insider’s perspective on how Bridget™ is fundamentally different from general-purpose generative AI tools.
Pioneering NVIDIA Nemotron for AI Output Validation
BridgeWise has become the first few companies in the world that adopted NVIDIA Nemotron multimodal models via Amazon Web Services (AWS) to independently validate AI answers before they reach users, an approach suited to regulated markets.
“In finance, accuracy and trust are everything,” said Elad Nachmias, CTO at BridgeWise. “Nemotron on AWS boosts our capabilities to rigorously test our AI outputs, checking for correctness, completeness, and any hallucinations, so that every release continues to reflect our deep domain knowledge and commitment to delivering reliable investment intelligence.”
This approach reinforces BridgeWise’s robust proprietary AI platform and its multi-layered accuracy framework as well as reflects a growing industry shift toward AI systems that can verify and critique their own outputs – an essential requirement for regulated financial environments.
Building Enterprise-Ready AI for Investments
Nachmias highlighted 3 foundational principles that guided the development of Bridget™ as a reliable, enterprise-grade chatbot.
- First, BridgeWise places strong emphasis on choosing the right model for final answer orchestration. “Not all large language models (LLM) are strict enough in following instructions,” Nachmias explained. For financial use cases, the answering model must reliably adhere to predefined rules, constraints, and compliance requirements.
- Second, instead of relying solely on traditional retrieval-augmented generation (RAG), BridgeWise developed a proprietary mechanism to detect and supply only the most relevant data and endpoints needed to answer a query. By limiting the information fed into the model, Bridget™ significantly reduces the risk of the model generating unsupported or fabricated responses.
- Third, robust infrastructure. Bridget™ currently handles over 30,000 investment-related questions per day, with infrastructure designed for large-scale institutional use. The platform is built to support a capacity of up to thousands of questions per minute. To maintain this scale and ensure low latency, BridgeWise leverages 14 different AWS regions to perform load balancing based on the fastest available model at any given hour.
AWS: Secure AI Infrastructure
AWS highlighted BridgeWise as an example of secure, enterprise-ready AI, where customer data remains protected and compliant with financial regulations.
“Amazon Bedrock provides a secure, regulation-compliant foundation where customer data always remains within the customer’s Virtual Private Cloud (VPC) and is never shared with third-party model providers to train or improve the underlying foundational models. Built on this trusted environment, BridgeWise is one of the most promising go-to platforms for AI-powered financial analysis,” said Yaniv Gueta, Principal Solutions Architect – Financial Services at Amazon Web Services (AWS), in the webinar.
Product Roadmap: From Assets to Entire Portfolios
Looking ahead, BridgeWise plans to expand Bridget™ from analysing individual assets to full portfolio analysis, allowing users to upload reports such as PDFs or tabular data for AI-powered insights. By combining its domain-specialized AI with the newest open reasoning and multimodal models, BridgeWise continues to advance the frontier of trusted, scalable investment intelligence. The new feature is expected to launch later this year, starting with a major international bank.
