In this case study:
Industry: Technology and Professional Services
Products and Services:
Azure Foundry
Azure Document Intelligence
Microsoft Purview Information Protection
Python
Location: USA
The Customer
A technology consulting firm that grew through acquisition: buy the customers, the revenue, the team.
What's less obvious going in is that every deal also brings a contract portfolio built by someone else's legal team, on someone else's templates, in someone else's systems.
After several acquisitions, that added up to hundreds of Master Services Agreements executed under different legal entities, none of them speaking the same language.
The Challenge: A Problem Almost Every Acquirer Recognizes
Contract chaos after a merger isn't an edge case. It's one of the most common and most underestimated integration failures:
- M&A failure rates run 70%-90% industry-wide (CapLinked)
- Poor contract management alone can cost up to 9% of annual revenue in leakage from missed obligations and unclear entitlements (Sirion)
- Technology and systems integration consistently ranks among the top - close problems, driven largely by fragmented contract records where each acquired entity keeps its own lists, naming conventions, and definition of "active". Gartner even reports that 83% of industry data migration projects fail outright. (Gartner)
This client lived that pattern for years. By 2026, its MSAs were consolidated into a single SharePoint library, a mix of signed PDFs, scanned copies, Word originals, and files protected under Microsoft Purview Information Protection. Consolidation solved storage.
It did nothing for holistic visibility and understanding, which is the part that determines whether a deal delivers value.
Basic questions had no fast answer:
- Which clients, across which acquired entities, had granted rights to use their name and logo in marketing?
- Which MSAs had quietly expired under an inactivity clause nobody was tracking?
- Which contracts had no matching account in Dynamics 365 because the legal entity name never matched the CRM record?
- Where are the gaps between the contract, the pipeline, and the invoice history?
The default fallback, manual review, is exactly why timelines slip. The Legal Executive Institute puts complex contract review at 5-10 hours each. Across hundreds of inherited MSAs, a full manual pass would run well past 1,000 hours of legal and analyst time, before accounting for chasing file versions, decrypting protected documents, or reconciling account names by hand. In practice, that's a multi-month, multi-analyst project competing for the same budget and attention as every other post-close priority.
The client needed a partner who could turn a static, multi-entity document library into one trustworthy dataset, fast, and without asking them to rip out or migrate anything first.
The Partner They Needed: Working with What Was Already There
Quisitive started by mapping the client's existing environment rather than proposing a new one. Contracts lived in SharePoint. Accounts lived in Dynamics 365 Sales. Revenue history lived in the data warehouse. The job was to connect those systems intelligently, not replace them.
The Approach
- Ran structured discovery sessions with the client's data and operations teams to learn which SharePoint fields were populated and trustworthy, which CRM fields sales and marketing relied on day-to-day, and where tribal knowledge was quietly filling gaps no system captured
- Worked with the client's security team to access IRM-protected files using the identity and permission model already in place, so no existing governance had to change
- Agreed field by field, in advance, on exactly which source system would answer each question (an MSA start date, an end date, a marketing usage flag) and what the rule should be for edge cases, such as treating a blank end date as no end date rather than as missing data
- Built a data dictionary documenting, for every output field, exactly how it was derived: pulled from SharePoint metadata, inferred by the AI model, calculated through Python matching logic, or sourced natively from Dynamics 365
That upfront alignment meant no ambiguity later about what a number meant or where it came from, and it meant the client's own team could validate results against fields they already recognized instead of trusting a black box.
The Solution: An AI-Powered Contract Intelligence Pipeline
1. Harvest
A custom harvester built on the Microsoft Graph API walked the SharePoint library, pulled every document tagged as a Master Services Agreement, and captured full metadata (created date, modified date, custom columns, sensitivity labels). Files under Microsoft Purview Information Protection were decrypted through the MIP SDK using existing access rights.
2. Read
Scanned PDFs went through OCR; native PDFs and Word files skipped straight to text extraction. Every contract went to an Azure Foundry hosted LLM model with a strict rule: no summarizing, return structured JSON with the exact clause text, a confidence score, and reasoning. The model answered two questions per contract: does it grant marketing rights to the client's name and logo, and does it contain an inactivity clause that could trigger automatic termination, including the period and what it measures from.
3. Reconcile
This is where post-merger contract work usually breaks down. Contract names rarely matched CRM records exactly, a direct result of onboarding multiple entities with different legal names and abbreviations. A straight name match resolved only about 55 percent of contracts to a CRM account. A tiered fuzzy-matching layer in Python, tuned separately for recall and precision and backed by a small curated alias list, brought that close to 100 percent.
4. Enrich
The pipeline joined extracted inactivity clauses with invoice history and Dynamics 365 opportunity data, applying a priority order (opportunity end date, then last invoice date, then contract effective date) to calculate an estimated termination date for every contract.
5. Publish
Results landed in one workbook: a summary view, a full contract detail sheet, a list of medium-confidence matches flagged for review, and the data dictionary. A separate extract fed validated results directly back into Dynamics 365.
The Results: Post-Acquisition Contract Debt, Resolved in Under a Week
- Identified contracts that had already terminated under their own inactivity clauses, and others at risk within the next twelve months
- Surfaced active sales opportunities sitting on top of already-expired agreements, a gap invisible without joining contract, CRM, and revenue data
- Gave marketing a source-cited answer on exactly which clients they can name and feature, across every acquired brand
- Gave sales visibility into which live pursuits need fresh paper before a deal stalls
- Gave legal a clear view of which agreements need renewal before they lapse
Work that would have consumed a team of analysts for months, and typically gets deprioritized during integration for exactly that reason, was done in under a week.
The Results: Post-Acquisition Contract Debt, Resolved in Under a Week
The same architecture extends to Statements of Work, NDAs, data processing agreements, vendor contracts, renewal notices, and compliance attestations - the same categories of documents that pile up after every deal.
Because every stage is versioned code rather than a one-off prompt, it reruns on demand as new contracts are signed and new acquisitions close.
For a company that acquires regularly, this stops being a one-time cleanup and becomes a repeatable capability: the same playbook, every deal, instead of starting from scratch each time.
Quisitive Is Your Partner for Bringing AI Into How Your Business Runs
We build AI-powered solutions on Microsoft Azure, SharePoint, and Dynamics 365 that turn scattered business records into governed, queryable data. For companies that grow through acquisition, this isn't a one-time fix. It's infrastructure for every future deal.
