Key takeaways
- Every early access member gets a recorded 1:1 Meta and Google Ads audit with the CEO, yours to keep regardless of what you decide about the platform.
- The Business Context Engine is live today: it connects ad accounts, Slack, Notion, Sheets, email, meetings and CRM in roughly four minutes, read-only.
- The roadmap covers more source integrations, an expanded automation library and deeper compounding of test history, but timelines can change.
- Context is classified by relevance, recency and business impact before any agent sees it, so more sources produce sharper outputs, not noisier ones.
- Your data is never used to train models, is isolated per client at the data layer, and is fully exportable or deletable at any time.
What early access actually means
Early access to Almond is not a beta label on a finished product. It is a structured first cohort where the team is still actively shipping, and where the people who join now help shape what gets prioritized next. That is worth being direct about before anything else.
Every member of the early access cohort gets two things immediately: access to the Almond platform as it exists today, and a detailed 1:1 audit of their Meta and Google Ads accounts with Almond's CEO. The audit is recorded, and the recording is yours to keep.
What the audit covers
The audit is not a sales call with a screen share. It is a working session focused on your actual accounts.
The CEO reviews your Meta and Google Ads setups in detail before the call. During the session, the focus is on what the account structure reveals: where budget is likely leaking, which campaign or ad set patterns suggest untested hypotheses, where the account history shows tests that were stopped before reaching significance, and where the current structure may be misaligned with how the business has changed.
Because the session is recorded, you can share it with your team, use it as a reference when onboarding new hires, or revisit it six months later to track what changed. A verbal debrief that lives only in someone's memory is not useful. A recorded, structured review of your accounts is.
The waitlist requires a work email. That is not a filter for its own sake. The audit is only useful if the person receiving it has real access to the accounts and real authority to act on what comes out of it. A personal Gmail address does not tell the team enough to prepare a useful session.
What you get access to on the platform today
The core of what Almond ships today is the Business Context Engine, the agentic memory layer that reads your ad accounts, Slack, Notion, Sheets, email, meeting notes and CRM in read-only mode. Connection takes roughly four minutes.
Once connected, the engine begins capturing tests, results and decisions as they happen. It classifies each piece of context by relevance, recency and business impact, so a strategic decision from a leadership call is weighted differently from a throwaway Slack message. Nothing is entered manually. Nothing needs to be re-uploaded when something changes.
Built on top of the engine, three products are available to early access members:
- Canvas, the AI visual dashboard. Live boards of charts, tables and AI answers, starting from prebuilt audit and analysis templates. Shareable and exportable in a click.
- Workflow Builder, a visual automation builder with 30 prebuilt Google and Meta automations. Because every workflow runs on the context engine, it acts on judgment rather than triggers. It will not scale a campaign your team decided to sunset on Monday.
- Integrations, built on open protocols including MCP, so your context works inside Almond's tools or inside any AI client your team already uses.
What is coming next
A few things are worth naming honestly here. Timelines in an early stage company can shift. What follows reflects current priorities, not a guaranteed release schedule.
More source integrations
The current integrations cover the sources where most performance marketing context lives: ad platforms, Slack, Notion, Sheets, email, meetings and CRM. The next wave of integrations is focused on sources where context gets created but rarely captured, including additional analytics platforms and deeper CRM read layers. If your stack has a gap, the early access cohort is the right place to flag it.
Expanded automation library
The 30 prebuilt automations cover the most common Google and Meta workflows. The roadmap adds automations that address less common but high-impact scenarios, particularly around budget pacing decisions that depend on business context rather than platform signals alone. The goal is to get to a point where the workflow builder can act on a Monday morning decision by Monday afternoon, without anyone re-entering that decision into a separate system.
Deeper compounding across quarters
The context engine already versions and retains every test, result and decision. The next development focus is on how that retained knowledge surfaces at the right moment. The practical target: when you are about to run a test, the system tells you whether you have run a close variant before, what happened, and what the specific failure mode was. Not just a flag that the test exists in the archive, but a structured summary of what the result actually meant.
Bring your own LLM, further extended
Bring-your-own-LLM is available today. The roadmap extends this to make model switching more granular, so different workflows can run on different models depending on what your security team has approved for each use case.
Why the context engine matters more as more sources connect
A reasonable concern when connecting many data sources to an AI system is that more inputs produce noisier outputs. That concern is valid and it is something Almond is built specifically to address.
Every source that connects to the engine feeds into a classification layer before any agent sees it. Context is ranked by relevance, recency and business impact. A CEO strategy call outranks a throwaway Slack thread. A test result from last quarter outranks a brief that was uploaded and never updated. The engine is not a data warehouse with a chat interface on top. It is a prioritization layer.
The practical difference: instead of fourteen recommendations with no indication of which one matters this week, you get the three that matter this week and the reason each one made the list.
Security for early access members
Because early access members are connecting real accounts and real business data, the security posture is worth stating plainly.
Your context is never used to train Almond's models or any third-party model. It is never pooled across customers. Every context store is isolated per client and per team at the data layer, not enforced by prompt instructions. You can export or delete your data at any time. Everything is encrypted in transit and at rest, versioned, and auditable. If something goes wrong, you can answer the question that matters: what did the agent know, and when did it know it.
Who the first cohort is for
Early access is most useful for a specific kind of person: a performance marketer, media buyer, growth lead or agency founder who is already running significant spend on Google or Meta, who has felt the specific frustration of an AI tool that does not know what happened last quarter, and who has enough context in their accounts to make the audit genuinely useful.
It is not designed for someone who wants a finished, fully documented product with a support team and an SLA. That product is being built. The early access cohort is the group that helps determine what it looks like.
If you run an agency, the audit has an additional dimension. The session covers your accounts, but it also surfaces patterns that are useful for thinking about how your agency's accumulated knowledge could be systematized. Nine years of pattern recognition across forty accounts is your actual product. The audit is a starting point for understanding what it would take to make that visible and transferable.
Where Almond fits
Almond (almond.ai) positions itself as the world's first agentic brain for autonomous performance marketing. The Business Context Engine is the foundation. Canvas, Workflow Builder and Integrations are what it powers.
The early access cohort is the fastest way to understand whether the context engine solves the specific problem you have been working around: the AI that does not know what your team decided on Monday, the brief that was current in March, the test recommendation for something you killed six months ago.
The audit alone is worth the signup. A recorded, CEO-level review of your Meta and Google accounts, with no obligation attached, is a concrete deliverable regardless of what you decide about the platform afterward.
The waitlist is at almond.ai. Work email required.
Frequently asked questions
- What do you get when you join the Almond waitlist?
- Every early access member gets access to the Almond platform and a detailed 1:1 audit of their Meta and Google Ads accounts with the CEO. The audit is recorded and yours to keep. The waitlist requires a work email so the team can prepare a useful, account-specific session.
- What does the Almond ad account audit cover?
- The CEO reviews your Meta and Google Ads accounts before the call and focuses the session on budget allocation patterns, untested hypotheses, tests stopped before significance, and structural misalignments. The recording is yours to share with your team or revisit later.
- What is on the Almond product roadmap?
- Current priorities include additional source integrations, an expanded automation library beyond the current 30 Google and Meta workflows, deeper compounding of test history across quarters, and more granular bring-your-own-LLM options. Timelines can shift, as with any early stage company.
- How does Almond handle data security for early access members?
- Your context is never used to train models and never pooled across customers. Each client's context store is isolated at the data layer. Data is encrypted in transit and at rest, versioned, and auditable. You can export or delete your data at any time.
- Why is the Almond waitlist work email only?
- The 1:1 audit is only useful if the person receiving it has real access to the ad accounts and authority to act on findings. A work email gives the team enough context to prepare a relevant session. Personal email addresses do not provide that signal.
Almond 2.0 early access
Get a 1:1 Meta and Google audit with the CEO.
Every waitlist member gets a 1:1 audit of their ad accounts. Work email only.