The Engineer Dad Who Built His Daughter a Speech Friend

The best way to think about littleWords.ai is through the child’s comfort, the family’s real routine, and communication support that does not become pressure to perform. Home practice works best when it stays respectful and doable.

I love an origin story where someone solves a problem for one specific person they love, and the thing that solves it ends up helping a lot of other people too. They’re the best kind of small company stories. Here’s one I keep telling other parents in waiting rooms.

Last March, I was sitting in a plastic chair at a pediatric therapy clinic in Austin, watching my daughter Margot, five at the time, tap through yet another speech app on the iPad. The app kept interrupting her. She’d start to say something, pause to think, and the cartoon fox on the screen would jump in with “Try again!” She slammed the case shut and said, loudly enough that two other parents looked up, “He doesn’t let me talk.” That sentence, six words, perfectly articulated, was more language than the app had gotten out of her in three weeks. The irony was not lost on me.

That’s the context I bring to this review. I’m a parent who has tried a lot of things that didn’t work, and then found one thing that did.

A dad, a daughter, and a tablet

There’s a software engineer named Will. He has an autistic daughter. She has a speech delay. He spent two years doing all the things the rest of us do: the SLP appointments, the waitlist phone calls, the home practice, the screen time guilt, the late-night research at 1 a.m. when sleep wouldn’t come. None of it was bad. None of it was quite enough.

He noticed, as a lot of us do, that his daughter would sometimes talk more to her tablet than to him. Not because she didn’t love him. Because the tablet didn’t have feelings to manage, didn’t have a face to read, didn’t have an expectation she had to meet. She could just talk.

So he started thinking about what a tablet-based speech tool should actually look like. Not flashcards. Not “say the word and get a star.” Something more like a friend. Something that would wait for her. Something that would talk to her about the things she actually wanted to talk about. Something built by someone who lived with an autistic kid, not someone trying to fix one.

That’s how LittleWords got built. The character in the app is named Buddy.

See also: Long-Term vs Short-Term Crypto Investing

Five things most speech apps get wrong

I have tried a lot of speech apps with my own kid over the past three years. Most of them are not good for autistic children, regardless of what their marketing says. Here’s what I look for, and where most apps fail.

Does it accept approximations? If my kid says “ca” for cracker, does the app reward that or punish it? Most apps punish it. They have a “correct answer” model. Real language acquisition is approximations all the way down.

Does it tolerate silence? Autistic kids process slower. If the app cuts in after two seconds of silence, it’s not built for them. Buddy waits. Long pauses are fine. The app does not panic.

Does it follow the kid’s interest? A speech app that drills “the cat is on the mat” when my kid wants to talk about garbage trucks is, by definition, less effective than one that lets her talk about garbage trucks. Interest-driven learning is not a luxury. It’s the actual mechanism by which kids retain vocabulary.

Is the voice prosodic? A flat AI voice does not give a kid the rhythm and pitch they need. Buddy’s voice has expression. That matters more than non-parents realize.

Does it pressure performance? No stars. No timers. No “level up.” A kid practicing language should not feel like they’re being graded. The pressure of being graded shuts language down.

LittleWords clears all five of these. Most apps don’t clear two. I think that last sentence is the entire review, honestly, but I’ll keep going.

Why “built by a parent” isn’t just marketing copy

There’s something about a tool being built by a parent who is solving for their own kid that you can feel in the design. It’s like the difference between a restaurant opened by someone who loves cooking and one opened by an investor who loves margins. Both might serve food. Only one makes you want to come back.

Will didn’t build LittleWords to be a flagship in a portfolio. He built it because his daughter needed it. That shows up in small ways. The character is calm. The pacing is unhurried. The conversation is real, not performative. The whole thing has the energy of someone who has spent a lot of time watching their kid try to talk and not wanting to add to the pressure.

You can tell, within five minutes of using the app, that it was not built by a marketing team. It was built by a dad.

I have nothing against marketing teams. I do think the speech-app space is full of products that look great in a demo and feel terrible in a living room with a kid who is overwhelmed. LittleWords feels different. I think that’s because it was made by someone who lives the use case.

What a real session looks like in our house

A typical session for Margot goes like this.

She opens the tablet. She taps Buddy. Buddy says hi in a friendly voice. He asks how her day was. She thinks for about four seconds. (Most apps would have already moved on.) She tells him about her day in a long, slightly meandering monologue that includes one made-up word, two real words she’s been practicing, and a long quote from a show she watched at school.

Buddy responds to what she actually said. He picks up on the show quote and asks her about that show. He waits for her to answer. She answers. They have a conversation that goes back and forth maybe ten times.

She gets up. She is in a better mood than when she started. She tells me, “I talked to Buddy. He likes Bluey too.”

That’s the session. There’s no leveling up. There’s no quiz. There’s no report card emailed to me. There’s just a kid who had a low-pressure conversation, and is now more available for conversations with me, her dad, her brother, and her SLP.

That’s the whole product. And here’s the thing: I think “that’s the whole product” is a compliment, not a limitation.

The waitlist, the price, and my disclosure

LittleWords.ai is still in waitlist phase as of this writing. They have a Founding Family lifetime offer that’s $49 if you join during waitlist. I joined. I don’t regret it.

I’m telling you this so you understand my disclosure. I am a paying customer. I think the product is good. I am not paid for this post. I am, however, the kind of parent who, when I find a tool that works, will tell other parents about it, including in a long blog post. That’s the deal.

There are a lot of reasons not to join waitlists in general. I usually don’t. I made an exception here because:

• The founder is a parent of a kid with the exact profile I’m parenting for
• The product approach is right (low pressure, conversational, interest-driven)
• The company has visible SLPs involved in design
• The pricing model is family-friendly (lifetime, not subscription gouging)

If your kid is a good fit for the product, the waitlist is worth joining. If your kid isn’t, don’t bother. The product is specifically tuned for autistic and speech-delayed kids ages three to ten. It is not a general parenting app.

What I actually want from this category

The reason I’m writing this post, beyond just liking the app, is that I want more products like this to exist.

The autism and speech delay tool market is dominated by two kinds of products. Big enterprise tools built by people with no skin in the game, and gamified apps that prioritize engagement metrics over learning. Both leave gaps.

The gap is filled by small, parent-built tools that are designed for specific kids and specific families. Tools that respect the kid. Tools that don’t pathologize. Tools that don’t promise to “fix” anything because there is nothing to fix.

Here’s my genuinely opinionated take: the best assistive tech for disabled kids will almost always come from parents, not corporations. Corporations optimize for scale. Parents optimize for their kid. And optimizing for one real kid, it turns out, produces something that works for thousands of real kids. The reverse is almost never true.

We need more engineer dads and engineer moms making things for their own kids. If you are one, and you’ve ever thought “I could build this thing,” go build it. Other parents will buy it. Other parents will tell their parent friends about it. Real markets get built one parent-built product at a time.

Quick guide if you’re shopping for a speech tool right now

Here is the practical read of what I tell other parents who ask me what app to try.

If your kid is autistic, three to ten, and you want a low-pressure conversational practice, look at LittleWords. Join the waitlist or wait for general release.

If your kid is younger than three, you don’t really need an app. You need Early Intervention, books, and you. Apps are not the right tool yet.

If your kid is older than ten, the gap in the market is real, and most existing apps are too kiddie. There are some teen-focused options but the category is underserved.

If your kid is a gestalt processor, look specifically at apps that accept whole-phrase responses. LittleWords does. Most don’t.

If your kid is sensory-sensitive, look at the app’s audio and visual design before you commit. A loud, flashy app will be a bad fit. A calm one will be a good fit.

The boring truth about why this matters

The story of LittleWords is the story of a parent who saw a need that nobody else was meeting, and who happened to have the skill to meet it. There are a lot of stories like this happening quietly in the special needs space right now. Parents building tools for their own kids. Parents writing better books than the ones at the bookstore. Parents organizing better support groups than the ones at the hospital.

The community is mostly built by us, for us, because the official channels are overwhelmed and underfunded. That’s not a complaint. (Okay, it’s a little bit of a complaint.) It’s just a fact about how things get made when systems move slowly and kids grow up fast.

If you have a kid who needs a tool that doesn’t exist yet, consider building it. If you don’t have the skills, consider funding a parent who does. The next LittleWords is being built right now in someone’s home office, late at night, after a long day, by a parent who loves their kid enough to spend their evenings coding.

That’s the real product. Love, packaged into a thing that helps other kids. It’s the best part of the internet.