Casework Navigator: DIY Edition
OPulling together this newsletter is one of my favorite parts of running the Casework Navigator program — getting to monitor the situation and call it work is pretty cool. The sources I look at and the actual process of writing have changed through trial and error over the last few years, and I’m honored to have heard from many of you that it’s a useful resource.
So before our final final newsletter later this month, I wanted to take today to talk about my process — what to monitor to stay on top of agency program changes, how to make sense of it all, and how you might build something similar for your own office. Understanding when and how to look for agency information that will impact casework is an overlooked but critical skill.
One flag before we dive in: as AI tools have evolved, I’ve leaned into using them to cut down on the repetitive work of this newsletter. Some of the tools I use are not approved for House and Senate use — be sure to check with your team on your ground rules, and I’ll end with ideas for getting most of the same value with tools that are approved.
But first! One quick event flag: Tomorrow, June 17, at 1 PM ET, I’ll be joining Lawyers for Good Government for a webinar on managing casework related to canceled federal grants. Come join!
Anne Meeker
Senior Advisor
POPVOX Foundation
Defining What’s Important/Relevant
One of the most fun and hardest things about casework is that — as experienced caseworkers know — everything shows up here. To keep this newsletter from sprawling in every direction, I had to work to create clear definitions for what was and was not relevant.
Over the past few years, I’ve developed five major categories for determining whether an article is relevant and helpful:
Prepare: an agency update or story that seems likely to cause an uptick in casework demand. For example, I could make a pretty good guess that the IRS switching to paperless processing would create some challenges for people who previously relied on paper checks for their refunds, especially unbanked constituents.
Explain: a news story, often an “inside baseball” story about what’s happening at an agency, that will help caseworkers understand what’s causing constituent challenges with agencies that show up in casework. For example, an article about ALJ hiring backlogs at SSA might explain what’s causing SSDI/SSI appeals backlogs, and therefore SSA-related casework.
Proact: an agency update or program shift where offices may want to consider issuing a proactive PSA to help constituents avoid confusion. For example, USCIS has fallen into this bucket frequently over the past few years as the agency has increased application fees, changed expectations on location for green card applications, etc.
Watch: an ongoing trend that may cause casework demand or impact the quality of casework demand in the future. For example, the ongoing saga of the VA switching to a new Electronic Health Records System, or the Postal Service’s ongoing cash flow challenges — both may not immediately impact casework, or may impact different areas at different times, but worth keeping an eye on.
Meta: news about casework itself! This was always a super fun bucket, and one of the most idiosyncratic on sourcing — which we’ll cover below.
It was equally helpful to define what was not relevant: purely political stories and anything touching campaigns or elections. Having this as a clear mental heuristic — and eventually a formal filter — also helped keep me out of the rabbit hole.
Now here’s what’s actually in my inbox.
Agency Sources
90% of the time, agencies themselves are the most authoritative and reliable sources on program rules, updates, and changes. For each agency in your portfolio, make sure you’re subscribed to:
Congressional staff email distribution lists: not every agency has them, but for the ones that do, these are a gold mine — and surprisingly easy to miss!
Public press lists: these can be hit-or-miss on genuinely newsworthy public information, but also helpful in a meta sense for seeing what agencies want the press and public to know about their work.
Social media: Yep, most agencies are on X, and worth a follow. For big agencies like the Department of State, that may also include following Cabinet secretaries and their immediate direct reports, since they often get first dibs at announcing big changes, as well as associated high-ranking officials like Ambassadors.
However, as caseworkers know, the “official” Secretary-level communications shop doesn’t always tell the entire story — for that reason, I’d also highly recommend subscribing to email lists for agencies’ Inspectors General. For example, you don’t want to miss out on this absolutely bonkers VA OIG story about one VBA employee authorizing thousands of incorrect decisions!
Two other “official” but non-agency sources are worth your time as well: the Government Accountability Office (GAO), and the Administrative Conference of the US (ACUS). These both fall squarely in the “explain” category, and can be very helpful in understanding how an agency is set up and why certain cases are processed certain ways, as well as underlying processing vulnerabilities that sometimes lead to casework.
Last but not least, one major source of agency-side news is Congressional hearings: especially if your bosses are prepared to ask questions about the trends and issues your teams see in casework, these can be great sources of data and insight into what’s really going on at agencies. I’ve shared how I automated our team’s process of monitoring Congressional hearings on Instagram, and our fearless leader Marci Harris has some tips on using AI tools to quickly turn hearing transcripts into formatted casework-specific summary memos.
Media Sources
Agencies may be the most authoritative sources, but they’re not always the fastest, and rarely the most candid. For that, we rely on great journalism. Besides the major outlets, I subscribe to the daily newsletter for each of these:
Federal News Network, especially the Federal Drive with Terry Gerton
Politico
Route Fifty for news about state-administered federal programs like Medicaid and SNAP if your office handles them.
There are also lots of more specific deep-dive Substacks and other newsletters, depending on your office’s local media landscape and your portfolio. For example, AILA’s newsletter for Congressional staff is a great resource; your local social services or advocacy organizations — I’ve valued the Center for Community Solutions’ 5 Things for my area — may have some good stuff too.
For casework news specifically: to catch local stories that don’t make national outlets, I set up Google Alerts for “constituent services,” “constituent casework,” “Congressional casework,” “legislative casework,” and every combination of those words. Not perfect, but always interesting — and well worth experimenting with for office-specific areas of interest.
Making Sense of it All
Subscribing to all of these means a lot of email — 10 to 30 items per day, plus hearings, social media, and alerts. Here’s how I manage it.
First: get it out of your main inbox. Create a rule in Gmail or Outlook to filter these into their own folder and mark them as read. Better yet: create a shared email account your whole casework team can use for newsletters and press lists. Either option will save your sanity.
The workflow underneath this newsletter comes down to three steps:
scan (find potentially relevant items),
tag (decide what category they fall into and note what the blurb should say), and
draft (turn those notes into newsletter copy).
Over the past few years, I’ve used AI tools to handle more and more of the scanning and drafting on a schedule, without me having to be in the loop every time. This has meant building scheduled “tasks” in Claude Cowork that automatically:
Scan RSS feeds for media outlets on my “approved” list for casework-relevant titles and pasting them into a Slack channel
Use my browser to navigate to Committee websites to check for relevant hearings, capture transcripts, and post summaries of casework-relevant news in that same Slack channel
Based on my tags, draft starting-point blurbs for those news stories that meet my standards for length, link placement, and nonpartisan-ness
Pull those blurbs into a draft of the newsletter that I can double-check and edit, following my standard format for appropriate sections
Test newsletter links to make sure they work and they go to right right places
In practice, this means that AI tools save me a lot of time spent scanning news stories and copying + pasting links from one doc to another, while preserving the critical moment for my judgment and insight — deciding why a story is relevant and useful to caseworkers.
Here’s a quick video of what that looks like in practice, and I’m happy to share the keywords list and the automation spec with anyone who wants them — shoot me an email!
That implementation uses tools that are not approved for House and Senate use. But the most important work here is AI-agnostic: deciding what matters to you and your office, building your categories and exclusion criteria, and thinking about how you want to capture and share what you find. That framework is useful whether you’re running an automated pipeline or just cleaning your inbox.
For those who want to layer in AI, here are some ways to do it with approved tools like Copilot, Gemini, and ChatGPT:
Build a keywords list first. I used the published CaseCompass taxonomy as a starting point and iterated with an AI tool to refine it around the specific types of stories, updates, and program changes that actually show up in casework. This list is the foundation of everything else — whether you’re doing a manual inbox scan or using any AI tool to help filter. Even if you don’t use AI tools, having this keyword list on hand is still a helpful resource for training new staff and interns.
Build a reusable relevance-filter prompt. For me and my workflow, the five categories plus exclusion criteria were a complete classification system — figuring out the “why” behind what’s relevant and what you want your team to do with it is also a helpful exercise. Write them up as a prompt, save it somewhere easy to access, and paste article summaries into ChatGPT, Copilot, or Gemini to quickly decide if something’s worth including and what category it falls into, and to turn it into the right format to share with your team. This was the critical piece that unlocked the automation for me — without it, AI tools failed to really capture my specific casework perspective on why these were relevant.
Use AI to process your Google Alerts digest. Alerts come as daily email digests that are easy to skim past. Paste a digest or a newsletter into an approved AI tool with your keyword-filter prompt and ask it to flag and categorize anything worth reading — or write you a daily brief on what’s relevant.
Draft blurbs from your notes. Maybe you want to task one caseworker or an intern with writing your team’s own version of a Casework Navigator newsletter. Once you’ve flagged relevant stories and added notes on why they matter, paste both into an approved AI tool with some guidelines on your preferred tone and style. Even without automation, this cuts drafting time significantly. Your first draft is only as good as your notes, so keep them specific.
Questions? Want a brainstorming buddy on using AI tools to tailor your own situation-monitoring setup? I’m here to help! Drop me a line at anne@popvox.org, or check out our team’s full AI resources at popvox.org/ai.
Casework Navigator Newsletter
Get tools and tips to level up your casework, as well as the latest casework news and agency updates.
