Data lineage and governance specialist Solidatus has expanded its AI capabilities in a new release that introduces enhanced AI Assistant functionality, bring-your-own-LLM support and new integrations designed to help organisations accelerate governance and data lineage workflows.
In the summer, the company upgraded its service offerings with update 2026.3 the contains a set of capabilities that enable clients to leverage AI and agents in a way that is designed to boost efficiency and productivity when using the company’s suite of data tools.
The new functionality allows organisations to connect Solidatus to approved LLMs, maintain ownership of their AI model choices and accelerate governance, lineage and modelling tasks through AI-assisted workflows. At the same time, it solidifies the human-in-the-loop principle to ensure model-response accountability and trust, ensuring users retain oversight of agent activity and responsibility for outcomes.“This is a major step forward… there’s a brand new set of capabilities which weren’t there before,” Solidatus Chief Executive Philip Dutton told Data Management Insight. “Now, the tasks you can set agents, especially in the Solidatus environment where we’ve exposed all of our technical capabilities, all of our modelling capabilities, really move the dial significantly.”
Data Trust
The move is a logical step for Solidatus, whose lineage and governance platform provides firms’ data with traceability, auditability and transparency that is essential for ensuring information fed into LLMs is trusted. By introducing agentic capabilities into its governance platform, the company aims to help organisations complete governance and lineage activities more efficiently while maintaining transparency, auditability and human oversight.
The changes are intended to give clients greater control over how AI is deployed, managed and audited. It includes support for Model Context Protocol (MCP), allowing external AI tools and assistants to securely access governed lineage information within Solidatus.
It also introduces persistent AI Assistant sessions, enabling users to continue investigations and analysis across multiple sessions.
Dutton said that the new suite of capabilities provides the ability for clients to customise their workflows through agent “skills” that combine prompts, organisational policies and business context. That can be further facilitated by the integration of clients’ favoured models.
“LLMs have become fairly ubiquitous and they just sit as a background capability layer,” he said.
“And so clients are able to direct their agents specifically to particular actions, particular kinds of expectations in Solidatus. And those are able to be shared at either enterprise level, so everybody gets the same expectations of the agent, or they can be customised with specific requirements for specific departments across the organisation.”
Specific Outcomes
Solidatus is among a throng of AI-enabled data management service providers who are offering agentic processing because agents are capable of far more reasoning and can create multi-step plans for execution to achieve specific outcomes.The new approach is giving the company’s clients the ability to ingest unstructured data into their Solidatus environment and process it to their own specifications. That includes analysing Python code and extract, transform and load (ETL) procedures.
By integrating their own LLMs into their Solidatus environment, clients can also conduct analysis against their governance policies and standards, Dutton said.
“One of the big challenges for organisations is: are they compliant today against a particular regulation? And that would often take a team of people weeks to determine because there is so much volume of data that needs to be analysed. With the LLM, again, that becomes a simple, single prompt and away the agent goes and does the analysis.”
Financial Institutions
While Solidatus serves organisations across multiple industries, Dutton said the release had particular relevance for highly regulated sectors such as financial services, where governance, auditability and accountability are critical requirements for AI adoption.
Dutton said the new capabilities are particularly relevant for highly regulated industries, where any use of AI must be auditable, governed and subject to clear oversight.
“In financial services, the bar of expectations is higher than in other organisations because of their highly regulated nature; therefore, when we release capabilities, because of our customer set, those capabilities really have to be battle-hardened before they’re released,” he said.
“Everything is audited, everything is governed, everything has workflow, everything is really enterprise-grade in a way that if you’re focusing on SMEs, you don’t need to have that same level of rigour.”
The new offerings are contained within a separate module that clients can bolt onto their existing Solidatus provisions.
The capabilities are delivered as a separate module because many organisations maintain strict controls around the introduction of AI technologies.
Solidatus’ bring-your-own-LLM approach allows organisations to use approved models within their existing governance frameworks, reducing barriers to adoption and limiting concerns around data sharing.
Dutton said uptake had already been strong among existing customers, many of whom had been looking for ways to accelerate governance and lineage work while remaining within established AI governance frameworks.
Nevertheless, Dutton said the new capabilities would benefit all of its Solidatus clients, from the heaviest users of the platform to those who access its tools irregularly.
Looking Ahead
Dutton said future developments could see AI capabilities move beyond user-initiated workflows towards background monitoring and oversight functions, while retaining human accountability and review.
In the next few months, Solidatus is also looking to build on these capabilities with an agentic solution designed to analyse complex legacy systems and reduce the manual work involved in documenting them.
“This will unlock a lot of people’s very valuable time from doing fine-grain analysis and doing more oversight review and an attestation,” he said. “It will be a huge game changer and we’ve already got customers who are testing that with real, live code bases.”
Subscribe to our newsletter
