Target customer, product tradeoffs, and competitive landscape
Astrid targets the Head of Legal Operations (or Senior Counsel wearing an ops hat) at mid-market companies with a legal department of 5–15 people. These teams manage 1,000–3,000 active contracts across customers, vendors, and partners—enough volume that manual review is unsustainable, but not so large that they have budget for a full CLM deployment with a six-month implementation.
Legal ops regularly needs to answer questions that span the entire contract portfolio: "Which vendors have uncapped liability?" "How many contracts auto-renew in Q3?" "Do any DPAs restrict ML training?" Today, answering these questions means opening dozens of PDFs, reading each one, and building a spreadsheet by hand. It takes days. Astrid makes it take minutes.
The system architecture defines six distinct workflows. For the MVP, I built Ad-Hoc Query & Batch Diligence end-to-end and deferred the others. Here's why.
| Workflow | What it does | Why it was deferred (or selected) |
|---|---|---|
| Ad-Hoc Query & Batch DiligenceBuilt | User selects contracts + columns, gets a sortable table with per-cell confidence scores and source citations | Highest pain, broadest appeal. Every legal team needs portfolio-level term analysis. Validates the core extraction engine, trust model, and cross-document resolution—all of which the other workflows depend on. |
| Renewal Intelligence | Daily cron scans for upcoming renewals, pulls contracts, generates prep briefs for Sales | Requires a live calendar integration (Salesforce/HubSpot deal dates) and outbound notification system. High value but high integration surface area—not ideal for a standalone demo. |
| HubSpot Auto-Sync | Extracts terms from signed contracts, writes structured fields to HubSpot deal records | Depends on a live HubSpot sandbox with OAuth, deal objects, and custom properties. The integration is narrow (one CRM) and hard to demo without a real HubSpot account. |
| Ingestion & Risk | Watches cloud storage for new contracts, ingests, extracts, flags risk automatically | The ingestion pipeline was built (it's required for batch diligence). The proactive risk monitoring layer was deferred because it requires persistent background jobs and notification infrastructure. |
| Procurement Audit | Cross-references AP payments against contract index to find vendors with spend but no signed agreement | Requires an accounting system integration (QuickBooks/Xero) to pull payment data. Narrow use case (procurement-specific) and a different buyer persona than the core legal ops target. |
Batch diligence is the foundational capability. Every other workflow is a specialization of "extract terms from contracts and present them in a useful way." By building batch diligence first, I validated the extraction engine, confidence scoring, cross-document resolution, and the review table UX—all reusable infrastructure. The other workflows add triggers (cron, webhook, AP feed) and destinations (HubSpot, Slack, email) on top of that foundation.
It's also the most demo-friendly workflow. A user can upload contracts and see results in under 5 seconds, with no external integrations required. The review table—with expandable nested rows, confidence dots, exceptions filtering, and natural language chat—is the single most impressive thing to put in front of an interviewer.
The contract intelligence category is heating up. Two well-funded players are most relevant to Astrid's positioning.