Behavioral Decision Testing · Southeast Asia

Know how your audience
will respond before
you spend a peso.

Predikta simulates Filipino consumer response to campaigns, messages, and concepts — in days, at 10% of the cost of traditional research. Currently in soft launch — testing, fine-tuning, and building the commercial engine before full GTM. Built on published behavioral science.

88%
Survey-alignment accuracy
Soft
Launch phase — full GTM imminent
68.9M
Modeled Filipino population
Real
Revenue & pilots — pre-full-launch
Manila & Singapore · netopia.ai · predikta.ai
Globe / 917Ventures — Term Sheet Signed
01The Problem

Brands launch blind.
Research arrives too late.

Campaign decisions happen weekly. Traditional market research takes 3–4 weeks and costs ₱100–120K per study. By the time insight arrives, the budget is already spent.

Research cycles can't keep up with campaign cadence
Brand and agency teams iterate messaging weekly — sometimes daily. Traditional research takes 3–4 weeks minimum. By the time results arrive, the brief has changed and the insight is answering the wrong question.
4 wks
Average turnaround, traditional study
💸
Testing every concept is economically impossible
A single market research study in the Philippines runs ₱100–120K. Most brands can afford one or two studies per year — so the majority of campaign concepts go to market untested. The result is avoidable waste.
₱120K
Cost per study vs. ₱10–15K with Predikta
🌏
Global AI tools produce generic Filipino insight
Western psychographic models don't encode collectivism, family purchase influence, bayanihan values, or religiosity as a decision driver. Generic LLM prompts produce generic output — useless for brands that need culturally specific insight.
18mo
To build localized calibration — per market
📊
Analytics tell you what failed — after you've failed
Most marketing tools are retrospective. They explain performance after budget is spent. There are almost no tools built to answer the only question that matters before launch: how will this audience actually respond?
$0
Value of insight that arrives after the campaign has failed
02The Product

Simulate your audience.
Before you launch.

Predikta builds psychographic models of Filipino consumer segments and runs behavioral simulations against campaign inputs — results in hours, no fieldwork, no sample bias.

01
Submit your campaign concept or message
Upload a brief, creative direction, copy variants, or a concept deck. Predikta accepts raw inputs in any format.
02
Select your target segment
Choose from pre-built Filipino psychographic profiles or configure a custom segment using HEXACO and Schwartz Values parameters.
03
Run behavioral simulation
Predikta runs thousands of simulated responses calibrated against 68.9M profiled Filipino consumers.
04
Receive scores and directional insight
Predicted sentiment, segment-level response breakdowns, and message refinement recommendations — in hours, not weeks.
Turnaround time
vs. 3–4 weeks, traditional research
Hours
Cost per simulation
vs. ₱100–120K per traditional study
10%
Consumer profiles modeled
Nationally representative, census-scaled
68.9M
Survey-alignment accuracy
Lab-validated · arXiv 2505.22125v1
88%
03Commercial Proof

Real workflows.
Real clients. Before full launch.

During soft launch — before any scaled commercial outreach — all three paying clients have already returned for multiple simulations, integrating Predikta into recurring pre-launch decisions. That repeat-use signal, before a real GTM motion has even started, is the most meaningful indicator we have.

72%
predicted
75%
actual result
Δ 3 percentage points
The result that gives us confidence in the Q2 validation
In a consumer finance soft-launch pilot, Predikta predicted 72% positive campaign response before the campaign ran. The client's follow-up survey returned 75% positive. Three percentage points delta. Two additional validations have shown consistent directional alignment. The 200+ campaign study (Q2 2026) will test this at scale with peer-reviewed rigor. If it holds, this is a commercially defensible prediction engine.
n=3 commercial validations during soft launch. Directionally consistent with arXiv lab results. The Q2 study is the first statistically robust commercial test.
Consumer Finance
Retrospective Campaign Validation

Workflow: Used Predikta during soft launch to assess campaign decision impact before committing to retrospective performance comparison. Client provided historical campaign data for validation against simulation outputs.

Why it matters: A real enterprise validation workflow — not a demo. The client is actively comparing Predikta outputs against observed campaign performance to evaluate commercial fit.

Enterprise workflow active Stage 2 in progress
FMCG / Wellness Brand
Pre-Launch Creative Testing

Workflow: Predikta integrated into campaign process for audience insighting, tone testing, ad-copy iteration, and benchmarking before launch — with a defined A/B testing path post-deployment to validate predictions.

Why it matters: A live workflow with a defined validation loop — not a pilot without exit criteria. This is the recurring workflow model we're building toward at scale.

Live campaign workflow A/B path defined
What matters most: every paying client came back
Without a sales team. Without scaled outreach. Without a full GTM motion. During soft launch, all three paying clients returned for additional simulations — integrating Predikta into recurring decisions rather than treating it as a one-time test. When full commercial launch begins, we will be scaling a workflow that has shown early stickiness — not discovering whether one exists. One-time pilots don't build a business. Workflow replacement does.
04Scientific Foundation

Not prompt engineering.
A calibrated instrument.

Predikta's behavioral model integrates HEXACO personality science and Schwartz Values Theory — culturally recalibrated for Filipino psychology. Collectivism, family decision-making, religiosity as a value driver, and economic risk aversion are not parameters you can add to a prompt.

📄  arXiv:2505.22125v1 · Co-authored with University of the Philippines · Published 2025
95%
Trait Alignment
Psychographic model matches real human profiles across all measured personality dimensions
88%
Sentiment Accuracy
Consumer sentiment simulation vs. real survey outcomes across 12 consumer scenarios
What this validated — and what still needs to be proven
The arXiv study validates that our psychographic modeling can replicate survey-style sentiment with 88% accuracy at statistical significance. This is survey replication accuracy — it proves the methodology works. What we're now testing commercially is whether simulation outputs predict real-world campaign outcomes, not just survey sentiment. The Q2 2026 commercial validation study (200+ campaigns) addresses that directly. These are distinct questions, and we keep them distinct.
05Traction

Soft launch.
Already paying.

These numbers were generated before any scaled commercial outreach. Full GTM launches imminently — the pipeline being built now is significantly larger than what has converted to date, and we expect the pace to accelerate materially once the commercial engine is running.

🚦
Context: soft launch only — full GTM has not started
Commercial activity began in late February 2026 as a controlled soft launch — product testing, workflow fine-tuning, and early client validation before full go-to-market. No scaled outreach has happened yet. The paying clients, pilots, and pipeline below reflect what has been generated organically and through targeted introductions during this pre-launch phase. Full commercial launch is the first major milestone this round funds.
₱70K
Monthly recurring revenue
Soft launch · pre-GTM
3
Paying clients
All have returned for repeat use
5
Active enterprise pilots
Converting to paid imminently
30+
Product testers
Soft launch cohort
Paying clients
  • thynkertech — Repeat simulations, active workflow
  • UNBOX Philippines — Campaign testing integration
  • GlutaMAX — FMCG creative validation
Active pilots
  • Home Credit — Consumer finance validation
  • Globe — Strategic + telco campaigns
  • AdSpark — Agency workflow integration
  • Inquiro — Research firm partnership
  • BEVI — Brand campaign testing
Qualified pipeline
  • AXA · Metrobank · Century Pacific
  • Lenovo · HONOR · OPPO · Xiaomi
  • Riot Games · DDB Group
  • Kojie-san · Cyberzone · Mapúa
Leading indicators we watch
  • Repeat usage rate per account
  • Pilot-to-paid conversion timeline
  • Commercial validation results (Q2 2026)
  • ACV trajectory in enterprise accounts
06Why This Gets Harder to Copy

Four ways we may build
an advantage over time.

01 · DATA FLYWHEEL
Accuracy may compound with every simulation
Every prediction generates validation data when compared against real outcomes. If this feedback loop strengthens accuracy over time — more usage → better models → better predictions → more usage — it becomes harder for a later entrant to catch up. We're 18 months into building this cycle. Whether it compounds meaningfully is still being tested.
Testable hypothesis · In progress
02 · LOCALIZATION INFRASTRUCTURE
Localized infrastructure takes time to reproduce
Built on 70K collected behavioral profiles and nationally representative respondents via University of the Philippines — scaled to 68.9M via census integration. Reproducing this work in each new SEA market requires local data collection, academic partnership, cultural calibration, and validation studies. That takes 18–24 months per market and cannot be shortcut with off-the-shelf data.
Philippines complete · SEA expansion planned
03 · CULTURAL PSYCHOLOGY BARRIER
Cultural calibration is not prompt engineering
Collectivism, family purchase influence, religiosity as a value driver, risk aversion in economic uncertainty — these require structured recalibration, not clever prompting. Our HEXACO + Schwartz Values framework was built through multi-year academic collaboration and peer-reviewed validation. A generic model won't replicate this without the same investment of time and local expertise.
Published · arXiv 2505.22125v1
04 · STRATEGIC DISTRIBUTION
Strategic access that cold outreach can't replicate quickly
Globe is the Philippines' largest telco — 80M+ subscribers. 917Ventures is their startup arm with a portfolio of companies that are natural Predikta buyers. The signed term sheet gives us faster access to enterprise decision-makers, portfolio design partners, and distribution introductions. This advantage holds only if we convert that access into repeatable revenue — which is what this round is designed to do.
Term sheet signed · Follow-on option
07What Has To Be True

Three falsifiable conditions
for venture-scale outcomes.

We're not asking investors to believe in inevitability. We're asking them to back a team with the right credentials, a real early signal, and three specific falsifiable hypotheses — each of which we can test within 18 months.

1
Simulation must align with real-world campaign outcomes — not just survey sentiment
The 200+ campaign validation program (Q2 2026) will be the first peer-reviewed test of commercial prediction accuracy. The 72% → 75% fintech soft-launch result gives us confidence it will hold. If it does, this becomes a commercially defensible prediction engine. If it doesn't, we'll know before the next raise, and the category thesis weakens accordingly. This is testable and falsifiable.
Q2 2026 · 200+ campaign study in progress
2
Customers must use Predikta repeatedly, not just experimentally
All three paying soft-launch clients have already returned for multiple simulations. The pattern emerging: test once to validate, then integrate into recurring pre-launch decisions. Scaling from 3 to 25+ clients while maintaining that repeat-use pattern is the key commercial milestone of this raise. One-time pilots don't build a business. Workflow replacement does.
Early signal positive · 3/3 paying clients repeat · pre-full-launch
3
Localized models must outperform generic AI tools on Filipino consumer insight
If culturally-calibrated Philippine models consistently produce better predictions than generic LLM prompting, the localization investment creates defensible product value. If accuracy converges, the moat weakens. The lab validation (88% accuracy vs. generic baseline) suggests localization matters significantly. Commercial measurement is underway.
Lab-validated · Commercial measurement underway
08The Team

The unusual combination
this category demands.

Building behavioral decision infrastructure requires enterprise execution, behavioral science depth, and the credibility to attract academic and strategic partners simultaneously. This team has all three — and has done it before.

Axel Kornerup, MPA
CEO & Co-Founder
Built and sold netopia — the Philippines' largest internet café network. Part of the founding team at InterVenn Biosciences (AI liquid biopsy, $100M+ raised from Softbank, Genoa Ventures, Xeraya Capital). Founded four prior ventures across internet infrastructure, telco, gaming, and clean energy. University of the Philippines MPA. 26 years building companies in markets where infrastructure is the product.
netopia — acquired InterVenn — $100M+ 4 prior ventures UP MPA
Jason Albia, MS
CSO & CTO & Co-Founder
Computational scientist with 40+ peer-reviewed papers and ~$4M in AI R&D funding secured. Co-designed Predikta's psychographic simulation architecture — the same framework behind the arXiv publication showing 95% trait alignment and 88% sentiment accuracy. Degrees in Computational Science and Applied Physics. This is not a prompt wrapper. He built a calibrated behavioral instrument.
40+ papers ~$4M R&D funding arXiv co-author Computational Science
Aldo Carrascoso
InterVenn Biosciences
$1B+ outcomes
Fundraising · US Expansion
Jojo Flores
Plug and Play Ventures
$10B+ portfolio
Enterprise BD · VC Intros
Jane Walker
ex-PLDT / Singtel / San Miguel
SEA Partnerships · Enterprise
092026 Milestones

Conservative targets.
Falsifiable milestones.

This $1.5M round has one commercial job: prove the first wedge is a repeatable, venture-scale business — not a project, not a consultancy, but a product with recurring demand and a clear path to Series A.

Now · Q1–Q2 2026
Execute full commercial launch
  • Transition from soft launch to full GTM
  • Convert active pilots to paying accounts
  • Publish 200+ campaign validation (Q2)
  • Build repeatable outbound sales motion
Q3 2026
Prove repeatability
  • 10+ recurring enterprise accounts
  • Documented repeat usage across cohorts
  • Commercial validation data published
  • Self-serve features shipped
Q4 2026
Build the Series A case
  • 12+ recurring accounts established
  • SEA dataset expansion initiated
  • Series A conversation earned, not assumed
10The Ask

Raising to ignite
full commercial launch.

$1.5M
Seed Round · SAFE · $6M Valuation Cap
Instrument
SAFE
Valuation cap
$6M USD
Round size
$1.5M USD
Lead / anchor
Globe / 917Ventures (signed)
35%
Sales & customer success
30%
Product & engineering
20%
Dataset enrichment
15%
Validation & operations
What this round funds — and what success looks like in 18 months
  • Execute full commercial launch. Soft launch validated the workflow and generated early paying clients. This round funds the transition: scaled GTM, outbound sales motion, and closing the active pilot pipeline into recurring contracts.
  • Publish the 200+ campaign validation (Q2 2026). The first peer-reviewed test of commercial prediction accuracy. If it confirms directional alignment, this becomes the Series A story.
  • Build a repeatable sales motion. 2 account executives + 1 customer success hire, focused on mid-size agencies and brand marketing teams where buying pain is immediate and clear.
  • Demonstrate a repeatable revenue engine. With 12+ recurring accounts, documented repeat usage, and commercial validation published, the business is de-risked enough for a disciplined Series A conversation — on proof, not promise.
  • Allocate a material portion of the round to data enrichment and model productization: improve prediction accuracy, operationalize retraining, and build a measurable replication gap so the next round is earned on proof, not promise.
Predikta · Seed Round · March 2026

The window to build
this in Southeast Asia
is open right now.

The methodology is published. The soft-launch signal is real. The strategic partner is committed. The team has built and sold companies before. We are raising to find out whether what's working at three early clients scales to thirty. We believe it does. This round is how we prove it.