We design and deploy a fully custom animated 3D AI companion for your brand, platform, or product — your character, your voice, your personality, embedded on your property. Not a template. Not a chatbot widget. Yours.
This is a concierge build service, not instant SaaS. Custom AI companions commission at $800–$2,000+ in the open market, often without the conversational engine, memory, or hosting. We include all of that — discovery, design, build, and delivery. We take a limited number of builds at a time.
What's Included
We've already built the engine. Your companion inherits every capability — you just get to decide what it looks, sounds, and feels like.
A bespoke animated 3D character in VRM format, commissioned or curated from a licensed character artist with full commercial IP rights included. The character is yours — not licensed back to us, not shared, not a stock avatar.
Commercial IP — full rights transferWe configure a neural voice via Amazon Polly — selecting from 20+ available voices and tuning speech rate, pitch, and pacing to match your brand. Personality, tone, conversation style, and topic scope are defined together in a discovery session and baked into the system prompt.
Neural voice via AWS Polly — 20+ optionsYour companion remembers visitors across sessions — their name, preferences, what they talked about last time, and where they left off. Built on DynamoDB with semantic recall: the companion can reference past conversations naturally, not robotically.
Session + long-term memory — DynamoDBIdle life animations, emotion blends, and real-time lip-sync tied to the Polly audio stream — so the companion looks alive, not frozen. Photo mode lets users capture a still with their companion. All rendered in-browser via Three.js, no plugin required.
Real-time lip-sync — Three.js / VRMThe companion isn't passive. It greets returning visitors, notices when someone seems stuck, and initiates conversation — check-ins, suggestions, or just a hello. Engagement triggers are configurable per your use case: sales assist, community hosting, care prompts, or entertainment.
Idle greet, check-in, and assist triggersWe deliver an iframe-based embed snippet that drops into any site or app — no backend work required on your end. Hosting is included in the monthly fee: our AWS infrastructure, HTTPS, CDN delivery, and ongoing uptime monitoring. We handle the ops so you don't have to.
Iframe embed — drop in anywhereHow We Build It
Every companion build follows the same five-stage process. Typical timeline is 3–6 weeks from signed agreement to live deploy, depending on character complexity and revision rounds.
We align on your brand, audience, use case, and companion vision. You fill out a creative brief — we make sure we understand what "right" looks like before anything gets built.
We commission or curate a 3D VRM character that fits your brand direction. You review concepts, provide feedback, and sign off before we move to build. IP transfers to you on delivery.
We configure the conversation system: voice selection, tone, knowledge base, topics, memory scope, and proactive triggers. You review a scripted demo conversation before code ships.
We wire the character to our engine — lip-sync, animation, voice, memory, brain — and package it into an iframe embed. You get a staging URL to test before it goes live on your property.
We flip the switch, verify the embed on your live site, and hand off the ops dashboard. Hosting, uptime, and model updates are covered in the monthly fee — we keep it running.
Who Builds With Us
If your audience expects a personality — not a search box — a custom companion is the highest-leverage thing you can put on your property.
Deploy a virtual persona that talks to your audience 24/7 — handles fan Q&A, sells merch, and stays in character when you're asleep. Your face, your voice profile, your IP.
Turn a static mascot into a talking, remembering, proactively engaging brand ambassador on your site or app. Product guidance, loyalty conversations, checkout assist.
A warm, accessible companion for patient engagement, senior living, or mental wellness apps — built on HIPAA-eligible AWS infrastructure. BAA available on Enterprise.
See Companion Care programAI companion characters for social and entertainment apps — with memory, relationship depth, and genuine personality. We build to your story bible, not a generic chatbot template.
A concierge companion that knows your properties, answers guest questions, books appointments, and remembers returning visitors — embedded on your listing site or booking portal.
Replace your help widget with a companion that actually knows your product, guides users through onboarding, and escalates to a human when it needs to. Engagement that retains.
The companion at /companion/ is our own live demo — same engine, same voice stack, same memory system your brand companion will run on. Try it. Ask it something. See it remember you if you come back. That's what we're building for you.
Try Before You Commission
Pick a vibe, name your companion, choose your brand color, and hear how it would greet your audience. When you're done, hit "Get My Companion" — your design brief pre-fills the application so we hit the ground running.
Welcome back — it's so good to see you again. I'm Aria, your concierge companion. How can I make today easier for you?
Tell me about what you offer.
This is a design mockup — not an instant build. Your choices become your design brief, and we'll build the real thing in 3–6 weeks.
See the real engine live at /companion/Investment
Setup covers character commission, build, and delivery. Hosting covers our engine, AWS infrastructure, updates, and uptime. Both are fixed — the character commission cost is included in setup, not added on top.
Signature
one-time setup
+ $198 / month hosting
Standard character
Setup includes character commission cost. 3–4 week build timeline typical.
Brand Companion
one-time setup
+ $498 / month hosting
Bespoke character + full IP
Setup includes bespoke character commission. 4–6 week build timeline typical.
Enterprise
custom setup — scoped to project
Hosting quoted at scope
Multi-character / integrations
Enterprise is scoped per project. We'll send a custom proposal after the discovery call.
The AI at /companion/ runs on the exact same stack — voice, memory, animation — your companion will use. Try it before you commit.
Every companion runs on Amazon Bedrock, Polly, Lambda, and DynamoDB — the same services powering enterprise AI at scale. HIPAA-eligible on Enterprise.
The 3D character we commission for you comes with full commercial rights. The IP is yours on delivery — not licensed back, not shared, not held by the artist.
We review every application personally. Quality is why we cap volume. You'll know within two business days whether we're a fit — and what to expect, honestly.
Start Your Companion
We review every application personally. We take a limited number of builds at a time — tell us your vision and we'll confirm fit and timeline within two business days.
Not a checkout — this starts a real conversation. Fill in what you know; we'll work out the details on a short discovery call.
Prefer email? studio@boojee.estate
Common Questions
The setup fee covers the character commission (we work with a licensed 3D character artist — the commission cost is rolled in, not added on top), the build of the companion engine configured to your brand, voice and personality tuning, revision rounds, the embed package, and deployment to our hosting infrastructure. You are not paying the artist separately. IP transfers to you on delivery.
We handle the commission on your behalf. You'll provide a creative brief (style references, character direction, brand vibe), we'll source the right artist for the style, and you review and approve concepts before anything gets finalized. The character is built with a full commercial license transfer — it's your IP, not ours, and not the artist's after handoff.
Our companions use Amazon Bedrock for conversational reasoning — the same infrastructure the rest of our BOOJEE stack runs on. Bedrock is HIPAA-eligible when configured under AWS BAA, which matters for healthcare builds. We don't use third-party AI APIs that would require your user data to leave AWS. The model is configurable per tier — Enterprise builds can specify the model family.
Yes, with caveats. The companion is a web component — it embeds cleanly into any modern browser environment. For mobile apps, it works in a WebView (which covers most React Native and Flutter apps). A fully native iOS/Android SDK is not included in current tiers — that's a custom Enterprise engagement. Tell us your platform in the application and we'll scope it correctly.
Session memory (included in all tiers) means the companion remembers everything in the current conversation — it won't ask your name twice. Long-term memory (Brand and Enterprise) means the companion stores a semantic profile for each returning user: their name, preferences, topics discussed in past sessions, and relationship history. When they come back, the companion greets them by name and can reference past conversations naturally. All data lives in DynamoDB under your account's namespace — not shared across builds.
Neither purely. The build is fully managed by us — you don't touch code. After launch, you get an ops dashboard where you can view conversation logs, see engagement metrics, and submit feedback for tuning. If the companion needs a significant personality update or new capability, that's a change request (typically included for the first 60 days on Brand and Enterprise). Routine model updates and infrastructure are covered in hosting — you don't manage a server.
We say "HIPAA-eligible" deliberately. The underlying AWS services — Bedrock, Polly, Lambda, DynamoDB — are covered under Amazon's HIPAA BAA. We execute a separate BAA with your organization on Enterprise builds. Full HIPAA compliance requires policies, training, and workflows on your side too — we can't claim that for you. What we can do is build on the right infrastructure, execute the BAA, and work with your compliance team. We do not claim the product alone is "HIPAA certified."
Signature builds typically run 3–4 weeks. Brand builds run 4–6 weeks. Enterprise is scoped per project but rarely under 6 weeks for multi-character or custom-integration builds. The limiting factor is usually character design iteration — the more specific your brief on day one, the faster revisions go. We build to a staging URL first so you can test before anything goes live on your property.