# The evidence behind AI-assisted software · GRID · Dscvry > GRID connects AI activity, software changes and validation results, and shows what is known about the work and where the record is incomplete. Source: https://dscvryai.com/technology/products/grid/ # The evidence behind AI-assisted software. What the AI did. What changed. What was checked, and what’s missing. Why it matters · How it works · Wherever the work is. Hosted Private cloud On-premise Edge Air-gapped ## Managing AI in software delivery. Enterprises are investing in AI to improve software delivery. They also need to account for its cost, understand its effect on productivity and keep oversight of the work it does. In regulated businesses, that oversight has to support compliance reviews and decisions about operational risk. - ### Compliance & assurance Can the organisation show how AI-assisted software was produced and checked? - ### Cost What work is AI spending supporting? - ### Productivity Where is AI helping work progress, and where is effort consumed without a satisfactory result? - ### Operational risk Where does AI-assisted work need closer attention? - ### Oversight Where is visibility reliable across the software estate, and where does it break down? ## A passed check can hide a missing record. GRID shows what is known and where the record is incomplete, instead of silently forcing an answer. ### The build passed. The history didn’t. On brightpath, the check succeeded while the account of how the change was produced has a gap. The record names what to investigate. ### Two sessions, one change. When two AI sessions overlap and touch the same files, GRID leaves attribution ambiguous if the evidence can’t establish which one produced the change. ### Different problems, shown apart. Failed validation, uncertain attribution and missing evidence call for different responses, so they are never merged into one score. ## One evidence graph. Better decisions. GRID resolves who and what did the work, links it to the artefacts it changed and the controls around them, and correlates it with outcomes, with the basis for every link recorded. Existing tools keep managing code, running checks and enforcing controls. Sources Vendor telemetry, enterprise APIs, endpoint evidence, EDR/XDR, SCM and CI/CD, ITSM and incidents. Evidence model Identity resolution, session and agent lineage, artefact attribution, control context, outcome correlation, and confidence and provenance. Decisions AI change provenance, assurance, economics and executive queries. How sure each link is **Observed:** directly captured. **Deterministic:** exact system linkage. **Correlated:** strong cross-system match. **Inferred:** reasoned, with confidence shown. A missing record is distinct from a failed check, and an association with a change does not by itself establish the cause of an incident. ## Evidence for management decisions. A team may be using more AI while spending longer resolving failed checks. An application may pass its build while the history of how it changed is incomplete. Those relationships need examining before anyone draws conclusions. - ### Spend, alongside the work. Usage and cost next to execution context, so engineering and finance can tell a demanding task from a recurring workflow problem. - ### The effort behind delivery. Sessions, retries and failed checks examined together. Productivity is judged against your own baseline: GRID supplies the evidence, it doesn’t turn usage into a return automatically. - ### Exceptions, into existing reviews. The attention queue says why a record needs review. Engineering and assurance examine the same account; release decisions stay in your approval process. ## Collection with accountability. GRID separates metadata and locally derived information from detailed investigation evidence. - ### Scope decides collection. Deployment scope defines what is collected, who may inspect it and how long it is kept. - ### Raw material needs a purpose. Raw prompts and source material require an explicit purpose. Metadata can be sensitive too, and is treated that way. - ### Your tools stay in charge. Code, tests and controls stay where they are. GRID connects the evidence around them. [Engineering evidence](https://dscvryai.com/solutions/software-engineering/) ## Where it runs. The AI goes to where the work is, from hosted to air-gapped. ### A managed environment. Start with a hosted service when speed of adoption and managed operations are priorities. Agree data handling, access and service requirements as part of the engagement. Discuss hosted · ## Put them to work. GRID is built on the same tested foundations as the rest of what we build. ### Models Specialist models, evaluated on your data and run inside your boundary. Explore Models → https://dscvryai.com/technology/models/ ### Agents Agents that run on their own and hand over to a person at the decisions that matter. Explore Agents → https://dscvryai.com/technology/agents/ ### Products Ready-made products for recurring jobs, from documents and contracts to sustainability data. Explore Products → https://dscvryai.com/technology/products/ ### APIs Model, agent and data APIs that plug into the systems you already run. Explore APIs → https://dscvryai.com/technology/apis/ ### Data Data collected, extracted and governed, with every value traced to its source. Explore Data → https://dscvryai.com/technology/data/ ### Evaluations Tests built from your own cases, rerun every time a model changes, so you always know how well it does your work. Explore Evaluations → https://dscvryai.com/technology/evaluations/ ## Bring us the work where wrong is expensive. Start with the operational problem. We’ll work out the data, models and agents it needs, and where it should run. [Talk to us →](https://dscvryai.com/contact/) **London**Discovery AI Limited · United Kingdom **Hyderabad**ReDscvry Technology Private Limited · India **Chennai**ReDscvry Technology Private Limited · India