Legaliser, subsequently branded Agreements.AI, needed more than generated contract text. Shahzaib’s product work connected retrieval, drafting, clause analysis and review controls in a contract application.
A generated clause can read fluently while failing to fit the agreement. The product needed to help a reviewer inspect parties, structure, obligations and clauses requiring attention.
The engineering problem was the process around the model: retrieve relevant material, prepare the output, apply checks and give a person a clear route to review it.
The platform used a retrieval-augmented architecture to bring templates and reference material into drafting and analysis. A layered validation pipeline checked structure and applied semantic review before presenting the result.
Clause summaries and risk indicators helped the reviewer navigate an agreement. Cases the pipeline could not resolve were escalated to a person, keeping human judgment in the workflow.
Project materials showing the workflows and interfaces described in this case.
The delivered application connected contract generation and analysis to a validation and review process. It exposed the parts requiring judgment instead of treating generated text as an automatically accepted agreement.
The technology owner must maintain the retrieval sources, model configuration and review rules together. Model or source changes can alter the output without changing the screen. Versioning those inputs and keeping review criteria explicit makes the workflow more understandable to a replacement team.
Project contribution: Shahzaib. The scope described above identifies the work behind this case.
Editorial review: Ego Eimi · Updated September 6, 2026 · About the team
Your technology
Start with the business
Tell us which system matters, what has changed and who controls it. We will define a reviewable next step. Read Foundation and the commitments and evidence.
