Skip to content

Company profile · 3 min read

Bloom: the brand layer for agents

Bloom turns brand sources into a canonical system that agents and products can call through a web app, API or MCP to create on-brand assets.

What it does

Bloom is a brand layer that turns scattered brand evidence—websites, decks, Figma files, social, images and guides—into a canonical system agents and products can call. It can generate on-brand images, video, copy and campaigns through the web app, API or MCP. The YC profile frames the product as brand infrastructure; the current homepage shows the ingest, systemise and create loop.

The fit is a marketing team, agency or product builder that wants agents to produce on-brand work without repeatedly re-explaining voice, aesthetics and references. Bloom is also relevant to software teams embedding brand context into their own products.

Why I’d look closer

The product has a clear operational model. A brand is created from ordered sources, analyzed into a ready state, then used by an API or MCP generation call. The docs expose asynchronous brand readiness, image generation, OAuth/API authentication and a source-based onboarding flow. That is better than treating brand as a static PDF.

Pricing is public. The pricing page lists a Scale plan at $100/month with 500 credits/month, shared workspaces, pooled credits and unlimited team seats. The YC launch reports more than 175,000 on-brand assets generated; that is company-reported traction, not independent creative-quality proof.

The founder context is relevant. The YC profile describes Ray Fitzgerald as the founder of Lucid and a former Inscribe builder who is focused on product and organic growth.

What I’d ask

How are conflicting brand sources resolved, how are generated assets versioned, and what happens when a brand changes its logo, palette or voice? I’d onboard one brand from real source material, test API/MCP outputs against a human review set, and inspect data retention and image rights.

My editorial take

Bloom is a strong fit for teams making lots of agent-generated creative where consistency is the bottleneck. The canonical-brand thesis is practical. The product earns its place if it keeps agents on-brand without turning the source-ingestion process into another manual content project.

Quick facts

FieldSourced detail
Buyer fitMarketing teams, agencies, founders and products embedding brand context
Public pricingScale $100/month with 500 credits/month; team/volume options shown
InterfacesWeb app, REST API and MCP
Traction signalCompany reports 175,000+ assets generated

Sources checked

Checked 2026-09-19.

SourceUsed for
YC company profileProduct, founder and traction claim
Bloom homepageCurrent ingest/systemise/create workflow
Bloom pricingPublic plan and credit structure
Bloom API docsBrand readiness and generation workflow

Cohort context

Bloom is listed in Spring 2026. In our 2026-09-18 directory snapshot, 112 of 193 listed companies in that cohort have YC’s primary industry label B2B (58.0%). This is a current-directory comparison, not an original intake count or a performance ranking. Nine-cohort dataset.

Public website snapshot

Observed 2026-09-19 16:16 UTC in raw homepage HTML. This records visible metadata and advertised links, not agent execution or product quality.

SignalHomepage observation
Product description metadataObserved
Canonical linkObserved
H1 or H2 headingObserved
Typed structured dataObserved
Docs/developer linkObserved
Pricing linkObserved
llms.txt linkObserved
Markdown alternateNot observed in this response

Public observations · Collection method. Missing links here do not establish that a capability or file is absent elsewhere.

Ali Abouelatta

About the author

Cofounder of Lazyweb (a16z Speedrun). Previously grew subscriptions and built native ads at Duolingo. Writes First1000 (90,000+ subscribers).

I cofound Lazyweb, which publishes this site. This is an owner-written publication, not an independent testing organization. Research notes distinguish observations, sourced reporting and editorial judgment.

First1000X