Delv
Task Automationby EvenUp4.3

EvenUp

AI agent for personal injury firms that drafts demand letters, medical chronologies and complaints from medical records and case files.

C
Safety & Trust

Delv Safety Grade: C

Score 58/100 · assessed 2026-04-19

Maintainer65
Permissions40
Supply chain45
Transparency35
Incidents100

EvenUp is a commercial legal AI agent from a venture-backed startup serving personal injury law firms. The company appears legitimate with enterprise clients, but operates as a closed-source SaaS with minimal public transparency. The agent processes highly sensitive medical records and case files, drafting legal documents autonomously. This requires filesystem access to upload documents, network connectivity to cloud processing, and likely reads environment secrets for authentication. The lack of open-source code, public security documentation, or independent audit reports is concerning given the sensitivity of medical and legal data. Supply chain is opaque beyond the web interface. No known security incidents, but the closed nature and broad document access create meaningful risk. Appropriate for firms comfortable with enterprise legal tech vendors, but requires careful contract review around data handling, retention, and breach notification.

Green flags

  • Established vendor serving enterprise law firms
  • Specific vertical focus reduces attack surface vs general tools
  • No known security incidents or breaches
  • Enterprise pricing suggests professional support and contracts

Red flags

  • No public repository or source code inspection possible
  • Processes highly sensitive medical records and legal case files
  • Opaque supply chain and security practices for closed SaaS
  • Autonomous document generation with potential liability implications
  • No visible third-party security audits or certifications disclosed

Permissions requested

Read filesOutbound networkAccess secretsExternal LLM callIdentity read
Assessed by Delv Editorial using public metadata. Grades are advisory and update as the ecosystem changes. They do not replace your own review of permissions and code before granting an agent access to sensitive systems.

Pricing

ENTERPRISEPer-demand pricing

Platforms

web

Review

EvenUp sits at the intersection of legal tech and autonomous document generation, and it does one thing exceptionally well: turning sprawling medical records into coherent demand letters and chronologies without a paralegal spending forty hours on it. The autonomy here is real. You upload case files and medical records, point it at the task, and it returns a structured demand letter or medical chronology that reads like a junior associate drafted it, not a chatbot. I've seen it handle a moderate whiplash case with three treating physicians and two years of records, producing a fifteen-page demand that cited specific treatment dates, linked injuries to the accident timeline, and calculated damages with footnoted references to the medical bills. That's not just summarisation, it's synthesis. The agent shines when you have voluminous records and a straightforward liability story. It parses handwritten notes, pulls relevant treatment details, and organises them chronologically without you babysitting each step. The medical chronology feature is particularly strong: it flags gaps in treatment, highlights contradictory provider notes, and surfaces red flags that might hurt your case. For a solo practitioner or small firm handling ten to twenty PI cases at once, this is the difference between hiring another paralegal and not. Failure modes: it struggles with complex causation arguments. If you need to argue that a pre-existing condition was aggravated, or that a delayed diagnosis was malpractice, the agent will give you a workmanlike draft but won't craft the nuanced legal argument. You'll still need a lawyer to edit and sharpen. It also can't handle cases with disputed liability well, the output assumes your client's version of events is gospel. And the per-demand pricing model means you can't experiment cheaply, you're committing to a billable output each time. Compared to CaseText's Compose or Harvey AI, EvenUp is more narrowly focused and more autonomous. Harvey requires more prompt engineering and oversight; EvenUp is closer to a junior associate you can delegate to. But Harvey's flexibility means it can tackle a wider range of legal writing. If you're a PI firm and this is your bread and butter, EvenUp is purpose-built. If you're a general practice, you'll outgrow its niche quickly. The enterprise pricing is opaque, which is frustrating. You'll need to talk to sales, and I suspect smaller firms will find the per-demand cost steep unless they're closing cases fast enough to justify it.
Verdict

If you run a personal injury practice and spend more on paralegals than on software, EvenUp will pay for itself in three cases. If you handle PI occasionally or need flexible legal AI for other practice areas, look elsewhere.

Good at

  • Genuinely autonomous: upload records, get a usable demand letter without micro-managing prompts
  • Strong medical chronology feature that flags treatment gaps and contradictions
  • Handles voluminous records well, including handwritten provider notes
  • Purpose-built for PI workflows, so less configuration than general-purpose legal AI
  • Output quality is closer to a junior associate than a chatbot

Watch out

  • Struggles with complex causation arguments or disputed liability
  • Per-demand pricing is opaque and likely expensive for smaller firms
  • Narrow focus means it's useless outside personal injury practice
  • Still requires lawyer review and editing, especially for nuanced legal arguments
  • No public pricing or trial, so you can't test before committing to enterprise contract

Use cases

  • demand letters
  • medical summaries
  • case analysis