AI for pipeline integrity audits
International Oil and Gas Service Provider · AI audit and on-premises system specification
OpenKit audited the gap analysis in this firm’s pipeline integrity work and specified an AI system for it that runs on hardware the firm owns. The firm now has a straight answer on which parts a language model could take on and which stay with a senior consultant.
Overview
The firm provides pipeline integrity consultancy to operators worldwide, with audits that sit underneath capital decisions, insurance positions and regulatory submissions. The step that matters most is a senior consultant cross-referencing an operator’s safety documentation against licensed industry standards clause by clause, using a methodology the firm refined over decades of field work. It was slow, repetitive work that depended on engineers who were approaching retirement.
The firm’s project sponsor wanted to know whether AI could take part of that load. Whichever way the answer came out, it had to stand up to procurement, engineering and a regulator. OpenKit ran five interviews across the consultancy and engineering teams, including ex-operators who had built management systems from the customer side. It mapped the audit process from RFI to report, marking where a model could step in and where human judgement has to stay, then wrote a specification and a costed roadmap.
0
Hosted models in the recommended design. Audit content and the firm’s methodology stay on hardware it owns.
2
Stages with a gate between them, so evidence search proves itself before drafting is funded.
3
Hosting options compared, each with a total cost of ownership behind it.
5
Interviews across the audit lifecycle, each paired with an evidence request, so every claim in the report pointed at evidence OpenKit had seen.
Approach
What OpenKit specified
OpenKit recommended a system in two stages that runs a language model on hardware the firm owns. Evidence search goes first because it is the cheaper, safer half. It also puts the firm’s hardware through real workloads before anything riskier sits on top of it.
- Evidence search, where staff ask a question in plain language and get back the passages from the audit documents that answer it. It is designed so that assembling evidence no longer takes days.
- Drafted findings, where the system breaks the framework into testable criteria and drafts a finding for each one, citing the source pages it came from.
- Consultant review, where a consultant accepts, edits or rejects every draft. No draft can reach a customer-facing report without that mark, and every decision goes into a record that cannot be changed.
- A gate between the stages, so drafting starts only if evidence search hits the agreed accuracy and stability targets on historical audits.
Hosting
How the hosting was decided
OpenKit compared three hosting options across open-source and commercial models and judged them on how well each one handled regulatory text.
- Pure cloud, ruled out from the start because audit content cannot leave the firm’s environment.
- Fully on-premises, the recommended starting point. The firm’s existing hardware made it the cheapest route to a working prototype as well as the safest.
- Hybrid, held in reserve for production, provided any overflow follows the same data-handling rules.
Result
A plan the firm’s own team can build from
The specification is written in enough detail for the firm’s own team to start building from it. OpenKit also wrote a report, shared across the business, on connecting the firm’s data layers and sources so every department it interviewed can work from the same evidence.
Voice
We worked with OpenKit, as we wanted a company with domain expertise in LLMs to look at our strategy and test the concept. The study was focussed and really gave us a validation of our concept, a technical roadmap and prioritisation of the developments. The OpenKit team engaged well during the project, and through capture of workflows and interviews gained a good understanding of what we do. OpenKit provided a good technical study and exceptional value for money.
Controls
Certifications and data controls
OpenKit certifications
- ISO 27001
- ISO 9001, UKAS-accredited
- Cyber Essentials
Controls on this project
- UK data residency
- Data sovereignty for the firm
Other engagements
What could your business do with AI?
Find out with experts who become part of your team. We uncover opportunities, get ideas working, and help your people build on the results. We reply within one working day.
Get in touch AI Audit