For decades, stewards have researched grievances the hard way. You'd dig through stacks of arbitration case files, search through USPS manuals, flip through MRS documents, and piece together precedent manually. It was time-consuming and imperfect. You'd often miss relevant cases because you didn't know they existed. Or you'd spend hours searching for one specific case you vaguely remembered. AI-powered research tools have fundamentally changed that.

I built CREA Research because I experienced this frustration firsthand. After 20 years carrying mail and serving as a steward, I realized that the difference between winning and losing grievances often came down to whether you had quick access to relevant precedent. Too many stewards were losing cases they should have won simply because they didn't know about favorable arbitration decisions or didn't have time to research thoroughly.

Now, AI is transforming steward research. What used to take hours can be done in minutes. What was impossible to find is now instantly accessible. And stewards who leverage AI research tools have a real advantage in negotiations and arbitration.

The Problem with Traditional Grievance Research

Before we talk about solutions, let's be clear about the problem. Traditional grievance research is broken.

When I was starting as a steward, I'd get a grievance involving, say, Article 14 safety violations. I'd need to find arbitration cases where arbitrators ruled on similar safety issues. So I'd go through cases one by one, reading summaries, trying to find relevant decisions. If I was lucky, I'd find two or three relevant cases in a few hours. If I wasn't, I'd miss important precedent entirely.

The problem was:

• Time-consuming. Researching a single grievance could take 4-6 hours just to find precedent.
• Incomplete. There are thousands of arbitration cases in the NALC record — spanning from the 1970s through 2025 — plus 1,700+ MRS documents and 73 USPS manuals with over 18,000 searchable sections. No human can read through all of that for a single grievance.
• Difficult to find connections. You might not realize that a 1979 Gamser ruling on overtime equitability (NC-S-5426) directly supports your 2026 OTDL bypass grievance, or that a Mittenthal decision from the 1980s (C-06238) established the absolute 12/60-hour limits you're citing.
• Disorganized sources. Arbitration cases were scattered, USPS manuals were dense and hard to search, MRS documents weren't indexed well.

The result? Stewards often went into negotiations or arbitration without the strongest precedent, because they didn't know it existed or didn't have time to find it.

This disadvantage fell hardest on newer stewards and stewards in smaller branches who didn't have access to veteran stewards with decades of case knowledge in their heads.

How AI Research Tools Work

An AI-powered grievance research tool fundamentally changes this dynamic. Instead of manually reading cases, you describe your situation — the specific workplace issue you're facing — and the AI analyzes it against the full body of contract language, arbitration precedent, and handbook provisions to surface what's relevant.

Here's how CREA Research works: When you have a grievance about, say, improper overtime distribution, you describe the situation in plain language. CREA analyzes it against:

• Thousands of arbitration cases spanning 1970-2025 — including NALC-designated key decisions, with outcomes spanning discipline, attendance, contract interpretation, overtime/scheduling, seniority/bidding, and more
• 73 USPS manuals — from the ELM and M-39 to the M-41 and JCAM, totaling over 18,000 searchable sections
• 1,757 MRS documents — Step 4 settlements, MOUs, and policy letters spanning M-00001 through M-02014

The AI identifies the contract articles that apply, finds relevant arbitration precedent, and pulls the specific handbook sections that govern the situation. You're not browsing a database — you're getting an analysis tailored to your specific issue.

The AI understands context. It knows that "Article 14 safety violation" and "unsafe working conditions" are related. It knows that M-00859 is the foundational MRS document for 12/60-hour overtime limits. It knows that a Mittenthal ruling from the 1980s on overtime caps is still the controlling precedent in 2026.

Real-World Impact: How This Changes Grievances

Let me give you some concrete examples of what this looks like in practice.

Example 1: Overtime Violations

A carrier at your station was forced to work 13 hours on a Tuesday. Before AI tools, you'd need to find the relevant contract language, the right MRS documents, and supporting arbitration cases. You'd probably know Article 8 applies, but would you know that National Arbitrator Mittenthal ruled in C-06238 that the 12-hour limit is an absolute? Would you find M-00859 (the MOU establishing the remedy of 50% premium pay)? Would you know that National Arbitrator Snow confirmed in C-18926 that M-00859 governs the remedy calculation? And would you find that the 2023-2026 Nolan Award added Article 8.4.G, making the penalty 2.5 times the base rate?

With CREA, you describe the situation and get all of that surfaced in one analysis. The contract language, the MRS documents, the arbitration precedent, the current contract changes — connected and explained in context.

Example 2: Article 16 Procedure Violations

A carrier faces a 14-day suspension. You need to know: Did management follow the six just cause sub-questions from the JCAM? Did a higher-level manager concur under Article 16.8? Is management citing expired discipline in violation of 16.10? Is the penalty proportional?

CREA analyzes the situation against over 13,000 Article 16 cases in the database. It identifies which just cause sub-questions are most vulnerable, surfaces relevant procedural rulings, and finds cases with comparable facts and outcomes — like C-31850, where Arbitrator Donna Thomas overturned an indefinite suspension on Article 16.1 and Article 16.6 grounds. Instead of spending hours piecing this together, you have it in minutes.

Example 3: Route Inspection Disputes

Management conducts a route inspection and adjusts your route based on the count week data. You think the evaluation was unfair. CREA can surface the relevant M-39 provisions (Section 242.311 on office time evaluation, Section 242.322 on street time selection), the key arbitration precedent (National Arbitrator Aaron's ruling in C-03207 that management cannot reduce office time below standard, pre-arbitration settlement M-00304 establishing no set walking pace), and the Step 4 settlement M-00242 confirming that comfort stops cannot be deducted from street time.

The July through December 2024 Contract Talk columns covered route inspections in extensive detail. CREA connects that guidance to the specific handbook provisions and arbitration precedent that apply to your situation.

The Steward Advantage

Here's what's changed: stewards who use AI-powered research tools now have a meaningful advantage.

When you can quickly identify relevant precedent, you:

• Negotiate from strength. Walking into Step A and citing National Arbitrator Mittenthal by name and case number changes the dynamic immediately.
• Prepare faster. You can build a complete research package in an hour instead of spending 4-6 hours digging through files.
• Make better decisions. Should you push this grievance to arbitration? When you can see how similar cases have been decided — across thousands of decisions — you make decisions based on real data, not assumptions.
• Handle more grievances. Because research is faster, you can handle more cases without being overwhelmed.
• Level the playing field for newer stewards. A steward with two years of experience using AI research tools can find the same precedent that a 25-year veteran carries in their head.

What AI Can't Do

I want to be clear: AI research tools are not replacements for steward judgment. They're tools that amplify your effectiveness.

AI can:

• Analyze a situation against thousands of cases and handbook provisions
• Surface relevant precedent — including cases you didn't know existed
• Identify the contract articles and MRS documents that apply
• Find patterns across cases and arbitrator rulings

AI can't:

• Know your specific membership and relationships
• Understand local politics and history
• Make strategic decisions about whether to arbitrate
• Negotiate with management
• Make the judgment calls that come from years of on-the-ground steward experience

You still need experienced stewards making decisions. But now you need stewards who know how to leverage AI research tools to make better-informed decisions.

The combination of experienced steward judgment plus AI-powered research is unbeatable.

Why Stewards Should Adopt This Now

Some stewards are slow to adopt new tools. They say, "I've been doing this my way for 25 years. I don't need to change."

I understand that perspective. But here's the reality: management is getting more sophisticated. They're hiring better labor relations specialists. They're doing their own research. The gap between stewards with good information and stewards without it is widening.

If you're researching grievances the old way, you're at a disadvantage against management that's better-resourced and better-informed. AI research tools level the playing field.

Additionally, younger stewards expect tools that make their work easier. If you want to attract new stewards and build bench strength in your branch, you need to make steward work more manageable. AI research tools do that.

Lino Miranda built CREA Research because he believes every steward deserves the research capability to fight effectively, regardless of whether they're in a large branch with institutional knowledge or a small branch where veterans are retiring. AI makes that possible.

Getting Started

If you haven't tried AI-powered grievance research yet, start now. Take one of your active grievances, describe the situation, and compare what an AI tool surfaces against what you found through traditional research. You'll see the difference immediately.

Pay attention to:

• How much faster you get relevant contract provisions and precedent
• Whether the tool finds cases or MRS documents you didn't know about
• How the analysis connects contract language to specific arbitration rulings
• Whether it helps you build stronger contentions for your PS Form 8190

The future of steward work is here. The tools are available. The stewards who use them effectively will win more grievances and better protect their members.