About Simulatte

Why Decision
Engineering.

Most consumer companies are still making seven-figure decisions on focus groups, gut, and demographic personas that read like fan fiction. We've watched too many launches fail not because the work wasn't smart, but because the work was structured to validate the decision after it was made instead of testing it before. We built Simulatte to fix that.

FOUNDER · 01
Founder

I started Simulatte after watching strategy work be confidently wrong.

I'm Iqbal Ahmed. I started Simulatte after spending years watching strategy work — agency, consulting, in-house — produce decks full of confident insight that turned out to be confidently wrong. The pattern was always the same: the research didn't predict outcomes, it gave the team something to point at when the launch underperformed.

The tools we had were not designed to be calibrated. They were designed to feel rigorous.

Simulatte is the opposite. Every decision we run through it produces a forecast with a confidence interval, a scope statement that's honest about where the calibration holds and where it doesn't, and a hit-rate counter that compounds with every engagement. We'd rather be honest about what we don't know than impressive about what we do.

Iqbal writes about prediction, AI, and behavioural science at Déjà Vu →

What we build

Calibrated synthetic populations.
Decision-ready.

Simulatte builds calibrated synthetic populations — decision-ready agents structured with 133 correlated attributes, an immutable Core Memory, a volatile Working Memory, and a cognitive loop that perceives, reflects, and decides. Brands route consequential decisions through these populations before committing real budget: ad concepts, pricing, positioning, packaging, launches.

The output is a Decision Artifact — a recommendation, a forecast, a calibration card. The full mechanism lives at simulatte.io/method.

What we've shipped

A small body of work.
Audited.

In the last three months we've shipped:

01
Six research studies validated across US and India populations, audited against Pew Research Center ground truth.
02
Five client engagements across fintech, e-commerce, DTC, and FMCG.
03
Calibration sprints that took India accuracy from a 46.2% baseline to 85.3% — the first system to replicate India's political and cultural landscape at population scale.
04
The largest external benchmark of synthetic populations against frontier LLMs (5,878 SHA-256 verified API calls). Simulatte landed 4.9× closer to the human self-consistency ceiling than the average frontier model.

It's a small body of work for now. We'll publish the full calibration index publicly as it grows. The empty seats are the point — receipts you can audit beat numbers you can't.

Where we operate

Calibration is live.
Footprint expanding.

Calibration is live in the US (88.7% distribution accuracy on Pew benchmarks) and India (85.3% — first at population scale). We're expanding into EU and LATAM as engagements ship. Our buyers are challenger CPG brands in the US and India, and the independent strategists and consultancies who serve them.

SIMULATTE JOURNEY · TIMELINE
Founding to calibration index · 2024 → 2026
Live
Feb 2026
Set Up
Founding · Persona engine v1
Mar 2026
First Benchmark
Pew US · 88.7% accuracy
Apr 2026
First Design Partner
CPG · live engagement
May 2026
Calibration Index
Public · honest receipts
Q3 2026
Blind Replays
Outcome calibration ships
Three principles

How we work.

01
Behaviour, not opinion.
We model what populations will do, not what they say they'll do. Survey data is an input to calibration, not the output of a study.
02
Simulation, not speculation.
Every decision variant gets run through the population before it gets run in the market. Outcomes are probable, not guessed. Every result is traceable.
03
A discipline, not a deliverable.
We don't produce reports. We produce a Decision Engineering practice the organisation can operate repeatedly — and one that improves with every engagement.
Recruiting design partners.
Reach us at iqbal@simulatte.io
Run the experiment at app.simulatte.io Coming soon