New Issue IQ, the solutions vendor dedicated to modernising primary bond markets, has announced a strategic partnership with deeptech AI company Boltzbit, to optimise the processing of new bond deal information. The collaboration reportedly delivers a processing-time improvement of approximately 74% by automating workflows that have traditionally been manual and fragmented.
Through this integration, New Issue IQ utilises Boltzbit’s proprietary “Instant Learning” technology and custom large language models to aggregate and standardise deal data. Consequently, clients can access aggregated data, analytics, and investment insights via a centralised hub within approximately 90 seconds of a bond’s announcement.
The primary market has long been constrained by unstructured data delivered through “messy” formats like emails and chat messages, says John Tokarowski, Founder and CEO of New Issue IQ. “In primary markets, speed is everything,” he explains, in conversation with TradingTech Insight. “The challenge lies in extracting usable data from the messy, jumbled formats buried in email systems. Boltzbit accelerates this process, allowing us to move from generic, off-the-shelf models to a solution that is fine-tuned for this specific domain. It is faster, more accurate, and more predictable – and the results have been compelling.”
The partnership aims to shift the buy-side workflow from reactive monitoring to proactive analysis. By automating the data entry process, the system allows investment teams to focus on strategy rather than data assembly.
“Buy-side clients no longer need to monitor their inboxes in fear of missing a new issue,” says Tokarowski. “Instead, they can rely on a dashboard that presents the full universe, allowing them to screen specifically for their investment strategy. The system populates automatically from distribution lists – emails, and soon chat messages. This transforms the workflow. Teams can immediately collaborate on issuer history and investment decisions. Crucially, analysis that used to be delayed until the afternoon by manual data assembly can now begin the moment the security is parsed.”
Boltzbit, founded by former Google and Microsoft AI researchers, utilises a Generative AI approach known as “Instant Learning”. This architecture allows models to adapt live within a client’s environment, contrasting with traditional models that often require extensive pre-training cycles.
Dr Yichuan Zhang, Co-founder and CEO of Boltzbit, details how this approach lowers the barrier to entry for specialised financial data processing: “New Issue IQ operates in a highly specialised domain. While clients possess the data, they often lack the expertise to transform it into private language models. Boltzbit bridges this gap, guiding them from data preparation to the deployment of accurate, efficient models. Key to this is our ‘Instant Learning’ architecture, which allows the model to learn live within the environment. This removes the need to go though long, expensive pre-training cycles to achieve perfection before deployment. Instead, data collection and model updates occur in the same loop. You can start with a ‘good enough’ model and refine it in real-time, avoiding the massive upfront resource investment usually required for high-quality data collection.”
Beyond the immediate efficiencies in the primary market, the partners believe the speed of data extraction will have positive downstream effects on secondary market trading and liquidity discovery.
“From a market microstructure perspective, the speed of data extraction creates a vital new source for secondary market liquidity discovery,” notes Ivan Mihov, Chief Revenue Officer at Boltzbit. “By combining new issue data with secondary market activity, firms can evaluate counterparties more fundamentally – tracking the full data lifecycle from initiation to final execution. This comprehensive view feeds directly into business decision models. It is particularly critical when paired with axes and transactional data, as the issuance process has clear, high-impact correlations with secondary market performance.”
Looking ahead, New Issue IQ and Boltzbit plan to extend these capabilities across additional asset classes. The companies intend to apply this autonomous agent technology to transform workflows in other areas reliant on outdated systems, such as the loan market.
Subscribe to our newsletter
