Finding a Residential Housing for Autism: What 20 Years Taught Me

If you’ve ever sat at your kitchen table at midnight, scrolling through program websites for your profoundly autistic child, wondering which glossy photo is hiding something ugly behind it, you already know this struggle. My wife and I have been searching for the right residential program for our son for over 20 years and I still don’t have it fully figured out. We use AI constantly now to speed up that search and honestly, it helps. But it also lies to you in its own way, not on purpose, just by being confidently wrong. This post walks through exactly what I’ve learned about finding a residential program for a profoundly autistic child, what AI got right, what it completely missed and what only showing up in person will ever tell you.

Key Takeaways:

• AI can help you research residential programs for autism, but it can’t replace boots-on-the-ground visits. I ran the same prompt through Claude, ChatGPT and Gemini and got three completely different lists of recommended programs.
• Staff reviews on Glassdoor and job sites often reveal what marketing brochures hide overmedication, cover-ups of elopement incidents and inconsistent care from house to house.
• You have to visit more than once and at off hours, to see how a program for a profoundly autistic adult or child actually runs when no one’s trying to impress you.
• Perfection isn’t the goal, what matters is honesty, respect and whether your child seems safe, loved and like they’re living their own life not just being managed.

Why did three different AI tools give me three different answers?

This is where I want to start, because it surprised me. I took the exact same prompt and ran it through Claude, ChatGPT and Gemini on the same day. Claude came back with five programs in Texas and three more around the country. ChatGPT found eight programs in Texas and eight nationally but completely missed a program Claude ranked as a top pick, The Autism Treatment center in Dallas and San Antonio. Then Gemini gave me five in Texas and six nationally and when I compared its five Texas picks to Claude’s five, not one program matched.

Same prompt. Same day. Zero overlap. That tells you something important: AI research tools are a starting point, not a verdict. They’re good at surfacing information you might not think to search for yourself, but they’re not consistent enough to trust blindly when you’re making a decision this big.

What did AI actually find that was useful?

In spite of the inconsistency, all three tools surfaced real, useful red flags, some of which matched what we’d already learned the hard way over 20 years of doing this without AI. Here’s what showed up:

• Staff allegations of residents being mistreated or overmedicated to manage behavior
• Leadership reportedly pressuring staff to cover up an elopement incident
• A 40-year-old program that had its license revoked twice in a short window
• Lack of basic accreditation, like the Better Business Bureau, for long-running programs
• Fine-print policies allowing a program to reject a resident for aggressive behavior without disclosing that during recruitment
• Parent complaints about physical restraint overuse, including unexplained bite marks
• Inconsistent quality and energy from one group home to another within the same program
• Skin shock devices still being used in some programs to modify behavior something I genuinely could not believe was still happening
• Beautiful, well-maintained facilities where staffing was quietly left as the parent’s responsibility

On the positive side, AI was also good at surfacing things like room setup, whether bathrooms are private or shared and clinical leadership credentials BCBAs, RNs and other staff qualifications. That part of the research was handled well.

Why can’t AI replace visiting in person?

Because AI can’t smell the boredom in a room. It can’t watch a resident sit alone in the dark, unengaged, while advertised staffing ratios sit on paper somewhere. That’s exactly what we found on one unannounced weekend visit. Residents parked in front of one small TV while two staff members chatted off to the side. No online learning, no engagement, just what I’d call infinite boredom. It left us pretty depressed.

Then there was the opposite extreme: a beautiful rural Texas program with gorgeous grounds and staffing ratios close to one-on-one. And then I watched four staff members take a young man to the ground, mats underneath him, while he said “let me up” over and over. It happened again a couple hours later with a different resident. Maybe there was history there I didn’t know about. Maybe other interventions had already been tried. But it didn’t feel that way and that loss of dignity is something I never want for our son.

A third program looked incredible on paper and in person three visits in, we were genuinely excited and so were friends whose son moved into one of the brand-new group homes. He was kicked out within two weeks after one incident of aggression. Lawsuits followed from other families later on. None of that showed up in an AI report. It only showed up by staying close to the program over time.

What should I actually look for when visiting a residential program?

Visit more than once and visit during off-hours when you’ll see what I call the “B team:” the staff who aren’t expecting company. Talk to other parents that are already in the program and if you can, talk to staff about what it’s really like to work there. And be willing to tolerate the things that don’t matter so you can actually see the things that do.

A few dirty dishes in the sink and an unmade bed that’s not the story. The story is whether everyone’s outside enjoying each other’s company, whether your child is excited to show you something new they learned, whether there’s a genuine and upbeat vibe when no one’s performing for the parents in the room.

How do I know it’s the right fit for my profoundly autistic child?

Honestly, I don’t look for perfection anymore, because I don’t think it exists for this population. Mistakes happen. Clothes get dirty. Routines get disrupted. Staff have hard days, and our kids have hard days too. What I’m looking for is honesty, trust, respect and joy especially in the small moments nobody’s paying attention to.

When our son Michael comes home from his weekend respite program, he’s usually a mess of dirty clothes, half his stuff left behind, iPad missing. But he gives me that look, that “Hey dad, I’m here, don’t bug me, I’m 36 and I’ve got my own life” look. That’s what I want. Not a perfect place. A place where he’s safe, loved, respected and can be himself.

What You’ll Learn

• Why running the same AI prompt through Claude, ChatGPT and Gemini can produce completely different program recommendations
• The specific red flags staff reviews on Glassdoor and job sites can reveal about a program
• Why licensing revocations and missing accreditations matter more than they seem
• What to watch for during an unannounced visit versus a scheduled one
• How to tell the difference between a cosmetic problem and a real one
• Why visiting more than once and at different times of day is non-negotiable
• What “good enough” actually looks like when perfection isn’t realistic

Notable Quotes

“Same prompt, same time, different recommendations.” -Mike Carr (00:54)

“There were allegations from the staff of the residents being mistreated or overuse of medication to zone somebody out.” -Mike Carr (02:43)

“A policy of being able to reject any resident for overly aggressive behavior and not sharing that openly during the recruitment process.” -Mike Carr (03:36)

“Skin shock devices to modify behavior. This is evidently a practice that is still going on in some of these programs.” -Mike Carr (04:23)

“We noticed residents by themselves in a bedroom with no lights on, just fidgeting… obviously they weren’t engaged.” -Mike Carr (05:44)

“There were four staff members that took this young man down to the ground… I heard him saying, ‘Let me up, let me up, let me up.’ I’ll never forget that.” -Mike Carr (07:19)

“It’s not about what the numbers say on the website or what you read in the reviews that matters. It’s what you see when you’re there. It’s what you feel when you’re there.” -Mike Carr (11:57)

“Perfection is not the goal. I think perfection with this group is impossible.” -Mike Carr (11:57)

Resources & Links

• Glassdoor – useful for finding employee and former-employee reviews of residential programs, including complaints leadership may try to bury.
• Better Business Bureau (BBB) – a basic accreditation check that can reveal whether a long-running program has earned outside trust.
• State licensing agency records – check for license revocations or violations tied to a specific program before you ever schedule a visit.
• AI research tools (Claude, ChatGPT, Gemini) – helpful for a first pass of program names, staffing credentials and red flags, but cross-check results since they rarely agree.
• J13 – the weekend respite program Mike’s son Michael currently attends while the family continues searching for a full residential fit.

Chapters

• 00:06 Introduction: the 20-year search for a residential program
• 00:54 Running the same AI prompt through Claude, ChatGPT and Gemini
• 01:53 Comparing zero overlap between Claude’s and Gemini’s Texas picks
• 02:43 Red flags AI surfaced: overmedication, cover-ups, license revocations
• 04:23 Skin shock devices and staffing shifted onto parents
• 05:06 What AI got right: room setup, clinical credentials
• 05:44 First visit story: unannounced weekend, disengaged residents
• 07:19 Second visit story: physical restraint incident in rural Texas
• 08:10 Third visit story: the program that looked perfect, then wasn’t
• 11:05 What actually matters versus what doesn’t
• 12:44 Bringing Michael home from respite and what “good enough” looks like
• 14:21 What’s coming next: guardianship and another dad’s story

FAQ

---

Transcription:

Mike Carr (00:06): Welcome back this week to another episode of Autism Labs. I’m going to talk about finding an amazing residential option, a residential program for our profoundly autistic child or hopefully yours too. We’ve been looking for a while, for over 20 years. So I’m going to walk you through what we’ve learned to look for, how to look for it, what really matters and how to find it which, as surprising as it might be, is not just from AI. And we use AI all the time. It’s helpful, but it can also be frustrating.

For example, yesterday I took three of the biggies ChatGPT, Claude and Gemini and I gave them exactly the same prompt. I led them through the research the same exact way and I’ll share this prompt in the show notes if you want to see exactly what I did.

Mike Carr (00:54): And guess what happened? All three came up with radically different recommendations. Same prompt, same time, different recommendations. Claude found five programs in Texas and three programs elsewhere around the country. ChatGPT found eight in Texas and eight nationally, but it totally missed the top program Claude recommended for the autism treatment center in Dallas and San Antonio. I just missed it. Then Gemini did a reputation and review analysis and came back with five programs in Texas and six nationally. I compared Gemini’s five Texas programs to Claude’s five and not one matched. Five totally different programs within Texas from the same prompt.

Mike Carr (01:53): This was a bit frustrating, as I said. I’ll put links to the full reports in the show notes so you can see exactly what each tool found and recommended. In spite of the inconsistencies, there was real value. Some insights matched what we’d already learned after doing this for 20 years without AI. I want to share those because they’ll probably help you too.

One thing to check: what are employees and staff saying about working there? You can often find this on Glassdoor or other job sites look for both positive and negative comments. Some of what ChatGPT, Gemini and Claude found included staff allegations of residents being mistreated or overuse of medication to sedate someone who was having a difficult day.

Mike Carr (02:43): So you show up and your son or daughter isn’t acting like themselves, because they’ve been medicated over a behavioral issue. Not good. Or leadership pressuring staff to cover up an incident where a resident eloped and wasn’t found for hours, causing a panic. Leadership didn’t want that getting out.

Second: licensing agencies not approving a program that had been around for over 40 years and revoking its license twice in a short period. Third: lack of accreditation from agencies like the Better Business Bureau you’d expect an established program to have at least that.

Mike Carr (03:36): Fourth and this was buried in the fine print: some programs have a policy allowing them to reject a resident for overly aggressive behavior, without disclosing that openly during recruitment. Fifth: employees mentioning legitimate parent complaints about physical restraint overuse a parent finding bite marks on their child’s arm or back and wondering why it wasn’t reported. Or inconsistency between houses. One house is clean with an energized manager, while another house feels neglected, with a manager uninterested in talking to parents. Inconsistent care and inconsistent KPIs from house to house.

Mike Carr (04:23): Sixth and AI uncovered this, which I genuinely couldn’t believe skin shock devices to modify behavior are apparently still used in some programs. Or a program with beautiful facilities, gardens and brand-new buildings, but staffing is quietly left as the parent’s responsibility. Who’s actually going to care for your nonverbal, profoundly autistic child with pica behavior or seizures, in that gorgeous building, if staffing isn’t guaranteed?

On the positive side, AI did surface a lot of what you’re probably already looking for: what the room looks like, whether it’s private or has a shared bathroom, strong clinical leadership, BCBAs and RNs on staff and other credentials.

Mike Carr (05:06): So the reports again, linked in the show notes might be helpful and insightful. But there’s one thing AI can’t replace: relentless phone-call sleuthing and visiting, probably more than once. Let me share a few things we’ve uncovered during onsite visits over the years.

Mike Carr (05:44): First, a few years back, we visited on a weekend after telling them ahead of time we were coming. It had to be a weekend since it was out of town. They advertised strong staffing ratios and engaged staff. Walking down the hallway, we noticed residents alone in bedrooms with no lights on, just self-stimulating behavior, unengaged, with no one paying attention. Then we got to the main common room, where nearly everyone sat on folding chairs watching one small TV, while two staff members chatted off to the side. No engagement, no ongoing learning, just what felt like infinite boredom. It left us pretty depressed.

Mike Carr (06:26): Another visit, this one during the week, was at a beautiful rural Texas program a couple hours’ drive, spacious grounds, gardens, outdoor activities and staff spread across multiple buildings with ratios close to one-on-one. We were thinking, “this is fantastic.”

Mike Carr (07:19): Then I saw something I’ll never forget. A young resident was taken down to the ground by three staff members, with mats underneath and around him for protection, one holding his arm, one on top with a mat, one holding his legs and a fourth watching. I heard him saying, “Let me up, let me up, let me up.” It happened again a couple hours later with a different resident in a different room.

Mike Carr (08:10): Maybe it was justified. Maybe there was history I didn’t know and other interventions had already been tried. But it didn’t feel that way and that loss of dignity and respect is something I’d never want for our son.

We visited a third program, upbeat and impressive everyone engaged, staff enjoying their work, immaculate new buildings with plenty of technology. We visited three times total.

Mike Carr (09:06): On the second visit we talked with the executive director, equally excited about their plans. On the third, we attended their grand opening of a multimillion-dollar residential community with brand-new group homes, some allowing lower-support residents to live independently but close together, with staff able to watch over everyone from a shared common area. We talked to verbal community members and their parents, who were thrilled. Friends of ours whose son had been in the program for a while were excited about him living there full-time.

Mike Carr (09:54): He moved into one of the first group homes. Within a week or two, he had an incident of aggression and was kicked out permanently. All that time, money and hope invested and he was gone within weeks.

Mike Carr (10:17): Later we heard about one lawsuit from another parent, then another. So the bottom line: you have to visit more than once and keep your ear to the ground. Use AI, but also talk to other parents currently in or considering the program and to staff if you can, about what’s really going on. Visit during off-hours, when you’ll see the “B team” are they engaged? Are they with it?

Mike Carr (11:05): You also have to be tolerant of things that don’t really matter, as long as the things that do matter check out. Maybe there are dirty dishes in the sink, not great, but if everyone’s outside having a great time together, that matters more. Maybe your child’s room isn’t perfectly made, but they’re excited to show you their newest friend or activity because they genuinely enjoy being there.

Mike Carr (11:57): It’s not about what the website says or what the reviews claim it’s about what you see and feel when you’re there. Is there a happy, upbeat vibe? Do people seem safe and respected? Do they seem like they’re living their own life, not just being managed? Perfection isn’t the goal. I don’t think it’s even possible with this population.

Mistakes happen. Clothes get dirty. Routines get disrupted. Staff have hard days and our kids have hard days too. What matters is honesty, trust, respect and joy especially in the small moments when no one’s watching, no one’s trying to impress you and you catch something out of the corner of your eye that tells you they’re doing right by your child.

Mike Carr (12:44): When our son Michael comes home from a few days at the R&R program at J13 we don’t have full residential yet, just weekend respite he’s a mess. Dirty clothes, missing water bottles, half his stuff left behind, his iPad forgotten. But what I’m looking for is a smile or more often, that cool look like, “Hey dad, I’m here, don’t bug me, I’m 36 and I’ve got my own life and I’m okay.” That’s what I want to see.

Mike Carr (13:35): Not a perfect place, a place where our profoundly autistic son or yours, can be safe, loved, respected and themselves. A place with routines, relationships, activities and most importantly, a staff and team that genuinely enjoys being there with them. A place where they’re not just cared for, but where they can become more fully who they were meant to be.

This place is hard to find, but not impossible. The more you know what to look for, the more you visit, the more you ask hard questions, trust your eyes, trust what you feel and listen to other parents the better chance we all have of helping our children build a life beyond us. That’s the hope. That’s why we started J13. We’re not there yet, but we’re on our way.

Mike Carr (14:21): I was going to talk about guardianship today, but after more research I realized I’m not qualified to cover it properly. I’ve got a special needs and guardianship attorney joining the podcast in a couple weeks. She’s excellent and experienced and she’ll cover the financial side, special needs trusts and guardianship in detail. Next week, I hope to interview another dad with a profoundly autistic son to share more of his journey. See you next week take care of yourself and take care of your autistic son or daughter too.