octa squared: the infrastructure around the model

octa squared. The infrastructure around octa.

The model is one component. The infrastructure is what compounds. octa squared combines six algorithms running in parallel: online experimentation, ensemble signal ranking, multi-model rotation, sequence policy learning, model distillation, and the macro analysis loop.
01 / Model vs process

The model trains on a release cadence. The process sharpens every hour.

Most AI for sales pitches imply the model keeps learning in real time. It does not. Model weights change when training runs land. What changes continuously is the orchestration around the model.

The model Weekly to monthly

Two release beats. Continued pre-training plus long-horizon RL on the corpus.

The infrastructure Hourly

Experiment results, rankings, and routing decisions update continuously.

01

One model component

octa trains on a release cadence: continued pre-training plus long-horizon RL on the CIENCE and Tenbound corpus. Weights change weekly to monthly, not per campaign.

02

Six algorithms running in parallel

Online experimentation, ensemble signal ranking, multi-model rotation, sequence policy learning, model distillation, and the macro analysis loop.

03

Every campaign generates data

Every send, reply, meeting, landing page click, and deal stage transition becomes validated learning per segment, inside the customer's own workspace.

04

Your data stays yours

New customers' data does not train octa. A customer can opt in to train a model for its own workspace, and its validated learnings feed that model.

02 / The lineage

Great compounding systems were never one algorithm.

The systems that defined search, research, and game-playing were infrastructures combining many loops, with outcomes feeding back into the inputs.

  1. 01

    Modern web search

    Hundreds of weak signals are ranked continuously. No single signal decides. The ranking infrastructure, not any single signal, does the work.

  2. 02

    Self-play game agents

    Search proposes moves, self-play generates new training positions, and the network learns from outcomes. The combination is what learns.

  3. 03

    Online experimentation platforms

    Variants compete for traffic. Outcomes are measured. Winners get more allocation per segment. The platform itself is the learner.

  4. 04

    octa squared

    Six GTM algorithms run around the octa model. Each loop validates inputs for the others. Validated learnings stay in the customer's workspace.

03 / The six algorithms

Six loops running in parallel. Each one sharpens the others.

The combination is the point: online experimentation tunes variables, ensemble ranking decides priority, multi-model rotation picks output, sequence policy decides the next step, distillation drops cost, and macro analysis watches the system.

01

Online experimentation

graph8 runs continuous controlled experiments on campaign variables: subject lines, send times, cadence depth, channel mix, landing page layout, and voice openers.

02

Ensemble signal ranking

Accounts, contacts, intent signals, and sequence variants are scored by combining many weak signals into one ranked output. Outcomes re-weight the signals.

03

Multi-model rotation

Generator, critic, and editor roles rotate across models per task. The winning output ships. The log shows which model won.

04

Sequence policy learning

Each touch is a state. From each state there are many possible next actions. octa squared updates the state-to-action policy from real outcomes.

05

Model distillation

Frontier models generate canonical exemplars for hard GTM tasks. A retrieval pool serves those exemplars to cheaper open source students.

06

Macro analysis loop

Weekly and monthly reports span capacity, customer outcomes, algorithm performance, and segment shifts so humans can review and course-correct.

04 / The compounding

Each algorithm's output is another algorithm's input.

Six loops, but the loops are connected. Validated learnings from one loop become the starting material for another.

Online experimentation Ensemble ranking

Validated variable locks become new ranking signals.

Ensemble ranking Sequence policy

Ranked accounts and signals shape which next best actions get tried first.

Multi-model rotation Model distillation

Winning outputs become teacher exemplars in the retrieval pool.

Sequence policy Online experimentation

Sparse or low-confidence states become hypothesis candidates for the next experiment.

Model distillation Multi-model rotation

Cheaper students enter the rotation and shift the cost and quality frontier.

Macro analysis loop All five above

Weekly reports re-weight which algorithm gets called for which task class.

05 / Your own model

Opt in to train a model for your workspace.

octa is built on a decade of authorized campaign data from CIENCE and Tenbound. New customers' data does not train octa. A customer can opt in to train a model for its own workspace.

With the opt in, the workspace's winning exemplars, variable locks, sequence policies, and distilled pool feed that workspace's model.

Winning exemplars

Variable locks

Sequence policies

Distilled pool

Workspace model
06 / The feedback pipeline

Capture. Label. Validate. Train. Ship.

For a customer who opts in, the five steps that turn its campaigns into its own workspace model.

  1. Step 01

    Capture

    Every campaign outcome, winning variant, state-to-action transition, and distilled exemplar lands in the workspace corpus.

  2. Step 02

    Label

    The infrastructure tags segment, variable, channel, and intent without humans in the hot path. Humans review aggregates.

  3. Step 03

    Validate

    Held-out replay on octa Bench. If the new exemplar would have won historical campaigns, it survives.

  4. Step 04

    Train

    Surviving exemplars enter the next training pass for the workspace model.

  5. Step 05

    Ship

    New weights deploy to that workspace. The six algorithms now run with a sharper component. The next loop starts.

07 / In production

What octa squared is doing right now.

Live inside graph8 campaigns. Every loop runs on its own cadence. The combination is what compounds.

6 algorithms

Running inside graph8 campaigns today, each on its own cadence.

Hourly

The cadence where experiment results, rankings, and routing decisions update.

Weekly

The cadence where macro analysis re-tunes the infrastructure.

07 / Pricing

Free for verified companies. Pay for what the agents run.

Every plan has unlimited users. Credits pay for sending, calling and running the agents. Three plans.

Save 20% billed annually
For verified companies Free
Free forever
No credit card

Inbound, your first campaign and the AI front desk.

  • 5,000 starter credits as you get set up
  • 250 credits a month
  • Unlimited website visitors and form capture
  • AI web chat and AI receptionist
  • Send from your own business mailbox
  • 1 credit per contact reveal
Start Free
To scale outbound Team
$99 / mo
Billed monthly

Unlimited contact data and your own sending infrastructure.

  • 10,000 credits a month
  • Unlimited contact data, fair use
  • Mailboxes, domains and LinkedIn accounts
  • Dialer and outbound calling
  • Outside-provider enrichment
  • Export, API, MCP and CRM push
Start Team
For high volume and agencies Platform
$499 / mo
Billed monthly

Everything in Team, at the lowest credit price.

  • 75,000 credits a month
  • Everything in Team
  • Lowest credit price, $0.0067
  • Agency client workspaces
  • Higher fair-use and export limits
Start Platform

Start free with no credit card, or pick Team or Platform and pay monthly or annually at signup. You can change plans inside graph8 at any time, and your setup and data stay where they are.

Every plan
  • Unlimited users
  • Free search and browse of graph8 data
  • Getting started package: company research, brand kit, first landing page
  • Bring your own AI keys

Talk to us. Qualifying teams get 12 months of unlimited contact data, free. Talk to us

What the plans include

One platform, three plans

Every plan opens the same platform for the whole org. The plans differ in monthly credits, the price of an extra credit, contact data and sending infrastructure.

What the agents run

Paid in credits
  • Campaigns and sending infrastructure Mailboxes, domains, and deliverability run for you, end to end.
  • 5-channel sequencer Email, phone, LinkedIn, SMS, and WhatsApp in automated sequences.
  • Dialer and AI voice agents Reps call from the built-in dialer. AI voice answers inbound calls, makes outbound ones and books the meeting.
  • AI inbox Every inbound email, LinkedIn message and reply sorted, drafted and answered.
  • Meeting routing and booking Route leads to the right rep and book the meeting automatically.
  • CRM, pipeline, and quotes Deals, quotes, and each rep's daily list in My Desk.
  • Studio content engine Brand, newsletters, landing pages, ads, and campaign content.
  • AI agents and workflows Operator and Executor agents for RevOps, deal reviews, and SDR work.

Contact data

Unlimited on Team
  • Unlimited contact data on Team and Platform Reveals are unlimited for every user on Team and Platform, under fair use. On Free, each reveal costs 1 credit.
  • Export, API and MCP bulk Exports, API and MCP bulk pulls, CRM push and webhooks on Team and Platform, under fair use.
  • Buying signals Visitors, intent, hiring, and job-change signals on every account.
  • Waterfall enrichment Outside providers on Team and Platform, pay on match or bring your own keys.
  • Free search and browse Your whole team explores 700M+ contacts and 100M+ companies.

For the whole team

Always on
  • Unlimited users, no seat fees Invite everyone. Pricing never scales with headcount.
  • Sending infrastructure on paid plans Mailboxes, domains and phone lines at cost plus 30%, LinkedIn accounts at $10 a month, paid from your credits.
  • Built-in CDP and integrations Warehouse, 500+ connectors, two-way CRM sync included.
  • Monthly or annual Pay monthly and cancel anytime, or pay annually and save 20%.
  • Fair-use rate limits Limits apply on all plans. They guard against scraping, not real GTM work.

Credits pay for sending, calling, AI and enrichment. Monthly credits are used first, then starter credits.

Data tax ROI calculator
octa squared

Bring your motion. Watch the infrastructure spin.

See which variables lock, which model wins each task, which signal gets ranked up, which sequence policy fires, and what the next loop ships.

Start Free Read octa