How BHASM works

How the AI runs your retention.

Every morning it scores every customer, decides who to reach and who to leave alone, writes the outreach, and shows its work — every decision documented and explainable. Here’s exactly how.

Data flow Engine signals The engine Traceability
How a score is reached

Every score, traceable to its inputs.

One customer. Urgency 94. The contributing factors and their relative significance.

How this score was reached 94
Silence durationprimary
Personal cycle driftsignificant
World context (pre-season)moderate
Purchase velocity trendsupporting
Relationship balancestable
Every score shows its contributing factors. No black box.

Exact model weights, aggregation formulas, and proprietary calibration logic are protected intellectual property. This ensures the engine remains effective, continuously improving, and difficult to replicate. All decisions remain fully auditable for your compliance and review needs.

Seed & Growth Factor labels and relative significance visible in dashboard.
Scale & Performance Full audit log export via API. Every decision, every signal, every outcome — exportable for compliance teams and regulators.
Architecture blueprint

Where BHASM sits in your stack.

Three nodes. No middleware to host. Every event passes the timing layer before it leaves your domain — output channels never receive a message the rules have not cleared.

Data sources Your existing stack
· Shopify / WooCommerce, Razorpay / Stripe, GA4 · CRM — OAuth-based CRMs · Snowflake / Databricks warehouse pull (Scale+)
BHASM · Timing Engine The timing layer
· Urgency model — 0–100 score per customer · Sentiment safeguard · frequency cap · disruption-aware timing · 1 in 6 messages held before they fire — every decision logged
Output channels Cleaned triggers only
· Omnichannel — WhatsApp, SMS, Email via Resend / Twilio · Your ESP via API key · Webhook out + CRM write-back for downstream automation
Zero-code integration

Connected in 3 minutes. No SDK to install.

There is no engineering lift. The timing layer is fully managed. You authenticate via OAuth with your store and your communication gateway; BHASM's background signal scan maps your historical data instantly and the first brief generates within 4 minutes of import.

3 min
to first connection
Shopify OAuth, Razorpay key, GA4 service-account — no SDK, no npm install, no webhook URL to deploy.
4 min
to first brief
Historical orders scored against your industry's natural rhythm. 270 customers typical first scan. Brief lands in your dashboard automatically.
0
infrastructure to host
No middleware. No webhook receiver. No engineering ticket. The timing layer runs in the BHASM cloud and only emits cleaned triggers downstream.

For Scale and Performance: Snowflake or Databricks pull activates once credentials are connected. Enterprise webhooks accept API-key auth. Setup remains hours, not months — the chip is built in; credentials light it up.

What changes for your customers

Reach each one on the week they’re ready.

A customer who’s about to come back gets a message that helps them. A customer who isn’t gets the silence that keeps them open to you next time. Same calendar, fewer sends, more returns.

Primary signal
Silence duration
Days since this customer’s last meaningful interaction. The strongest input to readiness — measures how primed they are to open.
What this changes
The message fires on the day the customer is ready — not on day-of-week. The email arrives because they wanted it.
Significant signal
Cycle position
Where this customer sits in their personal buying rhythm — computed from their own history, not an industry default. Identifies the day their window opens.
What this changes
Same email, peak day. Worth more than every message this week combined.
Moderate signal
World context
Cultural calendar, weather pattern, economic pulse. Pre-Diwali fires sooner; post-festival fires later. The brand doesn’t watch the world — the engine does.
What this changes
Peak buying energy captured automatically. The send rides the wave.
Reading the room
Sentiment state
Reads recent opens, clicks, complaints, tickets, refunds. Chooses which kind of send fires — the right message for this customer this week, not the campaign default.
What this changes
Every send calibrated to where the relationship actually is. The customer gets a message that fits them.
Permanent
Hard stops
Payment failure, opt-out, active dispute. Three permanent reasons BHASM never contacts. Logged in every audit. Everything else fires on timing.
What this changes
Three reasons the engine never reaches out. Every other customer gets the right send at the right moment.
The architecture

Built on laws, not prompts.

Most AI tools ask a language model what to do. BHASM runs the math first — the language model only writes the message after the decision is made.

18
Buyer archetypes
Each archetype has a distinct trigger, response rhythm, silence pattern, and message style. Identified from purchase pattern — not demographics.
∎
Timing precision
Every brief runs through a deterministic engine. Every account, every run, traceable end-to-end. Not opinion — mathematics that fires the right send.
62
Industry buying rhythms
52 days is the natural rhythm for luxury fashion. For a D2C supplement, it is past the window. BHASM knows when each customer's window opens across 62 industries — and times every send accordingly.
C7
Scoring dimensions
Cycle position · temporal decay · behavioural enrichment · world signal · archetype classification · sentiment read · frequency baseline. Seven inputs combined into one urgency score — the day this customer is ready to receive.
First brief uses industry calibration — 62 verticals, 18 archetypes as priors. Real personalisation kicks in around month two as the engine learns your base. What it knows at month six is not available at week one. That gap is not purchasable.
Start now

See it run on your customers.

Seed tier is free forever. Connect in 3 minutes. First brief runs automatically.

See your industry →
Every send, explainable
When BHASM reaches a customer — or holds back — you can see what led there. No black box. Built for the founder who wants to understand and the team that has to review.
Acts on the day, not the calendar
Reach the customer the week they’re ready, not the week the campaign was scheduled. Early enough that the relationship is still warm. Late enough that the message lands when they want it.
ROI in pricing
5,000 customers: typically recovers $2,500–$5,000 in month 1. Growth: $299. Scale: $999.
Multi-region compliance
DPDP Act 2023 · GDPR · CCPA. Regionally isolated data processing. Reach customers on Email, SMS, WhatsApp — the channel they actually open. Calibrated for your industry from day one.
Glossary · hover or tap any term
Urgency score — 0–100 composite. Five inputs. Every score traceable.
1 in 6 held — messages drafted then held until the window opens.
18 archetypes — each customer classified on first purchase.
62 industries — pre-calibrated cycles and sensitivities.
Personal cycle — silence measured against their average, not yours.
World signals — news, weather, festivals, economic windows.
Multi-layer timing — structural gates every message passes.
14-day attribution — what counts as a real recovery.

Every feature, one click away.

Hero · live demo
The live engine, on loop.
Watch the engine score an active customer, run the timing check, draft the message, and validate the recovery — in eight seconds. Same engine the dashboard runs.
See the live card →
Try the engine
Audit your own list.
Drop in a CSV or connect your store and payment platform. The engine scores every account, names the reorder windows, and shows the revenue waiting in the silence — on your data, in eight minutes.
Run the audit →
Audience switcher
E-commerce, SaaS, B2B.
Toggle the framework and the entire homepage recalibrates — customer name, urgency, value, channel — so you see your exact unit economics, channels, and compliance layer.
Switch the view →
The Retention Paradox
Why blocking 16% of sends raises revenue
Inbox placement falls from 94% to 71% when you force the 4th sequence item. The frequency limit holds the message to insulate domain health — the relationship survives so you can reach them when the timing is right.
Read the math →
Industry timelines
industry-scenario, Day 1 → Day 29
Six industries, six end-to-end recovery timelines — the moment BHASM acted and the value protected.
Open the industry scenarios →
Pricing · autonomy ladder
Confidence-gated send authority
Seed: human approval. Growth: confidence-gated autonomous send. Scale: full autonomy within the timing engine. Performance: SLA-backed timing. Each tier names the autonomy contract, not the feature list. SaaS, Stripe, Snowflake, enterprise integrations included on Scale and above.
See the autonomy ladder →
2,847 $29 $249,572