Pico and LDA Technologies have announced an integration that will see Pico’s Corvil Analytics platform support LDA’s NeoTap X aggregation technology in the forthcoming Corvil Analytics 10.2 release. The integration allows Corvil to ingest NeoTap X metadata – high-precision timestamps, per-port sequencing and data integrity signals – directly into its analytics pipeline, giving firms a more verifiable view of network activity for latency analysis, market data reconstruction, operational investigations and AI-driven workflows.
NeoTap X is designed to support up to 400Gbps of aggregate bandwidth, embedding metadata into packet streams at the point of capture so that downstream platforms can detect aggregation-layer problems such as packet loss, verify ordering and flag data ingest quality issues that are otherwise difficult to reconstruct after the fact. In practical terms, the integration lets Corvil consume richer signals about the condition of the data it is processing, rather than inferring that condition from the packets alone – moving the evidence of what actually happened on the wire closer to a single, auditable record.The announcement points at a question firms tend to reach only after their models are already in production: whether the data arriving at AI-driven monitoring and decision-support systems can be trusted when something goes wrong. That question does not get answered in the model layer. It gets answered much further down, at the aggregation layer, before anything downstream ever sees the data.
Why the edge matters more once decisions are automated
When a human analyst investigates a latency spike or a disputed fill, a degree of ambiguity in the underlying capture is tolerable – the analyst can caveat it, cross-check it, or escalate. When that same investigation is increasingly handled by automated monitoring and AI-assisted decision-support, the tolerance narrows sharply. A model cannot reason its way around a gap it does not know exists. If metadata generated at the network edge is lost before it reaches the analytics layer, the downstream system inherits a blind spot it has no way of seeing.
“AI and automation are only as good as the data they operate on,” says Jarrod Yuster, Founder and CEO of Pico, who frames the integration as strengthening Corvil’s role as a trusted source of truth for trading, risk and operational teams. As that work moves from people to systems, the integrity of the capture and aggregation layer becomes a constraint on how far automation can responsibly be pushed.
Vahan Sardaryan, CEO at LDA Technologies, positions NeoTap X as introducing what the company describes as an industry-first level of visibility into packet flows. With 100-picosecond timestamping precision, the underlying capability is not in question; what the integration changes is whether that precision is preserved end to end, or degraded somewhere between the tap and the dashboard.
A continuation, not a departure
For Pico, the move fits a clear two-year arc. Corvil Analytics 10.0, launched in 2024, introduced the platform’s machine learning and AI capabilities. The Corvil 12000, launched in late 2025, was pitched explicitly around AI-ready network observability and continuous high-throughput capture. NeoTap X support is the supply-side complement to those releases: having built analytics and AI tooling that consume network data, Pico is now hardening the pipe that feeds them, extending Corvil’s reach to a next-generation aggregation layer.It also reflects a shift in how this corner of the market is positioning itself. Corvil has moved towards serving both electronic trading and enterprise observability from one dataset – the same record underpinning execution analytics, compliance and infrastructure monitoring. Keeping that record consistent across teams depends on the data holding up at the point of ingest, and demand for this kind of verifiable, AI-ready network data is growing across the sector, spanning infrastructure analytics providers and exchange-side telemetry alike.
The integration will be available in Corvil Analytics 10.2. For buy-side and sell-side firms weighing it up, the open questions are practical ones: how it performs consuming NeoTap X metadata inside a live trading estate, and what it changes about mean-time-to-resolution or the confidence with which an automated system can act on what it sees – the kind of evidence that will determine how central edge-level data integrity becomes to the way firms specify their trading infrastructure.
Subscribe to our newsletter
