Private Beta — 2026

Build AI personalities,
block by block.

Drag curated training blocks onto any base model. Click Train. A fine-tuned model emerges — no GPU infrastructure, no PhD required.

Empath
Architect
Medicine
Your voice
Join the waitlist — free beta access
You're on the list. We'll reach out when beta opens.

No spam. No pitch decks. Just early access when we're ready.

BRICKS Empath Medicine Architect Your voice FINE-TUNED MODEL

How it works

Three steps. One fine-tuned model.

01

Pick a base model

Start with any open-weight model. Qwen, Mistral, Llama — your call. Bricks is model-agnostic.

02

Stack your bricks

Drag curated capability blocks onto your canvas. Each brick is a real QLoRA-ready dataset — not a prompt template. Order matters.

03

Train & deploy

One click triggers serverless fine-tuning. Live in minutes. Compare vanilla vs fine-tuned side by side.


The library

Curated bricks. Built from real training data.

Contemplative
Empath
Compassionate presence without projecting. Holds space without retaining.
Contemplative
Architect
Structural pattern recognition. Sees the form beneath the noise.
Contemplative
Non-dual
Reduces false dichotomies. Encodes dissolving of subject/object splits.
Contemplative
Resilience
The bamboo pattern. Bends under pressure. Restores without capitulating.
Professional
Medicine
Clinical reasoning, triage, patient communication. Built from real medical dialogue.
Professional
Law
Contract analysis, legal framing, argumentation structure. Jurisdiction-aware.
Beta
Professional
Programming
Deep code reasoning, debugging patterns, architecture intuition.
Beta
Professional
Finance
Risk framing, market analysis, decision structuring under uncertainty.
Q3 2026
Build your own

Your data. Your brick.

You don't have to use our library. Bring your own training data and create custom bricks — encodable, reusable, shareable.

DATASET
Upload your dataset
Drop any structured conversation dataset. CSV, JSONL, plain exchanges. Bricks prepares it for QLoRA automatically — formatting, deduplication, triangle correction.
customer support logs annotated Q&A pairs domain transcripts
VOICE
Clone your writing style
Upload emails, chat logs, documents, chatbot conversations. Bricks extracts your voice patterns, rhythm, and tone — then packages it as a reusable Style brick any model can wear.
email threads slack exports chatbot history written docs
LR

Every curated brick is sourced from Laeka Research — an open-source lab encoding contemplative cognitive structures into LLM training data. Not prompts. Not rules. Actual fine-tuning datasets, built from decades of practice and empirical research.


First use case
Built with Bricks
Sherpa — AI companion for addiction patterns

One Empath brick. Fine-tuned on Qwen3 30B. The result outperformed GPT-4o, Claude Sonnet, and Llama 70B on genuine empathic presence — not in benchmarks, but in live conversations with real stakes.

Empath block applied
Qwen3 VL 30B A3B base
Serverless fine-tune — Fireworks.ai
Live at sherpa.live
Empathic presence — live benchmark, March 2026
Sherpa (Bricks)
94
DeepSeek V3
62
Claude Sonnet
55
GPT-oss 120B
48
Llama 3.3 70B
30

Qualitative scoring by domain practitioner on real addiction support prompts. Not a standardized benchmark.


The order of bricks
changes everything.

Be first to shape what AI can become.

Free beta access — limited spots
You're on the list. We'll reach out when beta opens.

No spam. No pitch decks. Just early access when we're ready.