OTACTA / THE COMPANY

Building self-calibrating decision systems.

Building self-calibrating decision systems.

Building self-calibrating decision systems.

Systems that make recurring business decisions, measure what happened, and recalibrate how the next decision gets made.

Systems that make recurring business decisions, measure what happened, and recalibrate how the next decision gets made.

Our first product: Sofia →

How it works

DECISION LEDGER

CYCLE

DECISION

OUTCOME

ATTRIBUTION

RECALIBRATION

C-0418

Raise price on SKU-2231 by 3%

Units −1.2%, margin +2.4 pts

71% price · 29% seasonality

Elasticity −0.41 → −0.38

C-0419

Shift $2.1k ad spend to marketplace

ROAS 3.1 → 3.6

64% channel · 36% promo

Channel weight +0.06

C-0420

Produce 1,800-unit batch, week 42

Fill rate 98.2%

82% demand plan

Batch rule confirmed

C-0421

Hold reorder after 3PL rate change

Storage cost −$640

90% 3PL rate

Cost floor updated

C-0422

Pause promo at margin floor

Margin +1.1 pts

58% promo · 42% mix

Promo threshold −0.5 pts

C-0423

Move 400 units to faster-selling channel

Sell-through +8 pts

Channel mix likely primary

Allocation weight +0.04

C-0424

Reduce week 43 batch after demand shift

Excess stock −12%

Demand signal · lead time

Batch floor adjusted

C-0425

Raise reorder point for top SKU

In-stock rate 96% → 98%

Forecast · supplier delay

Safety-stock rule updated

C-0426

Pause low-return creative spend

ROAS 2.4 → 2.9

Creative mix · timing

Spend threshold raised

C-0427

Shift promo to lower-margin risk set

Margin +0.8 pts

Offer depth · product mix

Promo guardrail tightened

C-0428

Advance replenishment by four days

Lost sales risk reduced

Demand uptick · lead time

Lead-time buffer revised

C-0429

Test 2% price change on selected SKUs

Units steady, margin +0.7 pts

Price change · seasonality

Elasticity estimate refined

C-0430

Rebalance marketplace budget

ROAS 3.0 → 3.3

Channel shift · demand

Channel weight +0.03

C-0431

Hold production for slower variant

Weeks of cover 9 → 7

Variant demand · inventory

Variant mix recalibrated

C-0432

Change reorder cadence for key supplier

Cash tied in stock −6%

Order frequency · lead time

Cadence rule updated

Evidence layer · Live Decision Model™

Evidence layer · Live Decision Model™

Every cycle is kept. Every next decision draws on all of them.

Every cycle is kept. Every next decision draws on all of them.

01

The category

A decision system that gets sharper with every cycle.

A decision system that gets sharper with every cycle.

Every business makes the same decisions again and again: what to stock, produce, spend and price. Software records the results. AI makes the work faster. Neither changes how the next decision is made because of what happened after the last one.

Every business makes the same decisions again and again: what to stock, produce, spend and price. Software records the results. AI makes the work faster. Neither changes how the next decision is made because of what happened after the last one.

Decides against an objective

Works within set boundaries toward a stated business target.

Measures what happened

Tracks the result of each decision and attributes it to its causes.

Recalibrates the method

Updates how the next decision is made from that evidence.

02

The architecture

Live Decision Model™

Live Decision Model™

The LDM is the persistent layer underneath every decision. Each decision is measured, its outcome attributed to what caused it, and the method recalibrated. Decisions and outcomes accumulate into a living ledger of evidence, so the system knows what worked, under which conditions, and why.

The LDM is the persistent layer underneath every decision. Each decision is measured, its outcome attributed to what caused it, and the method recalibrated. Decisions and outcomes accumulate into a living ledger of evidence, so the system knows what worked, under which conditions, and why.

Measurement

What changed after the decision.

Attribution

How much of that change the decision caused.

Recalibration

How the method changes for the next cycle.

03

First product

Sofia builds a live operating model around your objective.

Sofia builds a live operating model around your objective.

Sofia is the first product built on the LDM. Give it an objective and boundaries. Sofia models how demand, inventory, production, spend, margin and cash affect each other, makes the recurring decisions, and recalibrates from what happens. We are proving it first in lean consumer-product and consumer-health businesses.

Sofia is the first product built on the LDM. Give it an objective and boundaries. Sofia models how demand, inventory, production, spend, margin and cash affect each other, makes the recurring decisions, and recalibrates from what happens. We are proving it first in lean consumer-product and consumer-health businesses.

Sofia

Objective: “Improve EBITDA 5% next quarter.”

Objective: “Improve EBITDA 5% next quarter.”

Demand

variance 0.4%

Inventory

SKU-2231 · 21d cover

Production

wk 42 · 1,800u

Marketing spend

ROAS 3.6

Margin / P&L

+2.4 pts

Cash

WC −$12k

Meet Sofia →

13% → <0.5%

13% → <0.5%

Forecast variance

Demand forecasting · F&B producer

+$4,820

+$4,820

Margin recovered in 30 days

Inventory and margin · skincare brand

3.4×

3.4×

Return on Sofia fees

SKU pricing · B2B hardware distributor

WHY NOW

Agents can act. The next generation learns from what happened next.

Agents can act. The next generation learns from what happened next.

01

Systems of record

stored what happened.

02

Systems of intelligence

explained it.

03

Agents

can act.

04

Self-calibrating systems

learn from what happened next.

OTACTA

FOUNDER

FOUNDER

Usman spent a decade in commercial operations at Haleon, Allergan and Sanofi, building the pricing, demand and marketing-mix models behind the decisions those businesses ran on, and re-fitting them by hand, every cycle, as conditions moved.

Usman spent a decade in commercial operations at Haleon, Allergan and Sanofi, building the pricing, demand and marketing-mix models behind the decisions those businesses ran on, and re-fitting them by hand, every cycle, as conditions moved.

Usman spent a decade in commercial operations at Haleon, Allergan and Sanofi, building the pricing, demand and marketing-mix models behind the decisions those businesses ran on, and re-fitting them by hand, every cycle, as conditions moved.

That was the job: be the layer that recalibrates. Sofia and the Live Decision Model are that layer, built to run on its own.

That was the job: be the layer that recalibrates. Sofia and the Live Decision Model are that layer, built to run on its own.

That was the job: be the layer that recalibrates. Sofia and the Live Decision Model are that layer, built to run on its own.

Portrait of Usman Janvekar

Usman Janvekar

Founder & CEO

Connect on LinkedIn →

Investors and partners

Investors and partners

Otacta is early by design. We’re building it alongside a small group of investors, operators and strategic partners who want in at this stage. If that’s you, I’d like to hear from you.

Looking for Sofia? Customers start at otacta.ai/sofia.

usman@otacta.ai

Meet Sofia

Otacta’s Live Decision Model™: a living ledger of decisions and outcomes.

© 2026 Otacta Inc.