Behavioral Science Meets AI
AI simulations replace surveys and focus groups at 10x speed and 90% lower cost
Predikta is Southeast Asia's first commercially deployed behavioral simulation platform with peer-reviewed validation. We enable brands to test messaging, products, and campaigns against synthetic personas that behave like real consumers — before spending on production or media. Built on psychographic profiles modeled from 68.9M Filipinos using national datasets and validated behavioral research.
⚡
10x
Faster than traditional research
Results in hours vs. 3-4 weeks
💰
90%
Lower cost vs. surveys/focus groups
₱10-15K vs. ₱100-120K per study
🎯
±3%
Prediction accuracy in live pilots
Matched real survey results
Proof: Real Business Impact
Customer Example #1
Consumer Finance: Validating ₱2M Campaign Before Launch
Major SEA fintech company (5M+ users) · Loan product messaging validation
Challenge: Client had 3 competing campaign concepts for a new loan product targeting middle-income Filipinos. Traditional research (focus groups + survey) would cost ₱120K and take 4 weeks — too slow for agile campaign iteration.
Solution: Used Predikta to simulate 10,000 synthetic personas matching target demographics. Tested all 3 concepts for sentiment, trust, and intent to apply. Completed in 48 hours for ₱15K.
Validation: Predikta identified Concept B as strongest (72% positive sentiment). Client validated with small n=200 real survey — Predikta prediction matched within 3%. Campaign launched on schedule with ₱2M media spend.
₱2M
Campaign value de-risked
Customer Example #2
Beauty Brand: Message Testing Improved Campaign ROI 18%
Philippine beauty/wellness brand · Social media campaign optimization
Challenge: Client planning ₱800K social media campaign for new product launch. Wanted to test 5 different messaging angles but traditional A/B testing requires significant media spend upfront.
Solution: Simulated audience response to all 5 messaging variants using 5,000 synthetic personas matching target demographics (female, 18-35, Metro Manila, skincare interest). Predikta identified clear winner: "dermatologist-recommended" angle scored 68% positive sentiment vs. 45-52% for alternatives.
Outcome: Client launched with recommended messaging. Campaign delivered 3.2% conversion rate vs. historical 2.7% baseline — 18% improvement in campaign ROI. Client attributes lift to better message-market fit identified via simulation.
₱0
Media spend for testing
💰
Raising $1.5M Seed · $6M Valuation Cap
Strategic backing from Globe/917Ventures (term sheet signed, follow-on option). Current: ~$1.6K MRR / ~$19K ARR (₱90K/mo from 3 clients). Raising on $2M qualified pipeline (Home Credit, Globe, AXA + 5 active pilots) and founder track record (Axel: InterVenn founding team, $100M+ raised). Funds accelerate sales hiring (2 AEs + CS), Indonesia/Vietnam datasets, and self-serve product features. 18-month runway to $1.5M ARR, Series A ready.
Current Traction & Metrics
ACV Range
$200-1.1K
/month (₱10-60K PHP)
Based on 3 paying clients
Pilot → Paid
2/5
converted in 60 days
3 pilots still active
Sales Cycle
30-60d
first call → contract
Median 45 days
CAC / Payback
$630
~2 month payback
₱35K PHP, early data
Defensibility: Why This Is Hard to Replicate
Four Compounding Advantages
Not "we're early" — actual structural defensibility that improves over time
1. Dataset Moat
Ground Truth Behavioral Data
We own proprietary psychographic datasets for 68.9M Filipinos (modeled from national census, consumer studies, behavioral research). Built via UP partnerships — not available to competitors. Expanding to Indonesia (273M) and Vietnam (98M) creates geographic lock-in. This is our Bloomberg terminal equivalent.
2. Feedback Loop
Models Improve With Usage
Every pilot generates validation data — we compare predictions to real outcomes and retrain. Accuracy compounds: early movers get worse models. We're 18 months ahead on Philippine data. First-mover advantage = better training data = higher accuracy = more customers.
3. Localization Barrier
Cultural Nuance Hard to Replicate
Western tools (Qualtrics, Nielsen) lack local behavioral frameworks. Filipino decision-making differs from Western models (collectivism, family influence, risk aversion). Our HEXACO + Schwartz implementation is culturally calibrated. Replicating this requires 2+ years local research.
4. Strategic Distribution
Globe/917Ventures Partnership
Strategic shareholder (term sheet signed, follow-on SAFE option) provides: distribution to 80M+ Globe subscribers, portfolio ecosystem (Brave, Inquiro, AdSpark) as design partners + live customers, and data feedback loops that strengthen accuracy. Accelerates Philippine market dominance before competitors can establish foothold.
Paying Clients (3)
- thynkertech — AI product studio ($545/mo)
- UNBOX Philippines — Tech media ($273/mo)
- GlutaMAX — Beauty brand ($455/mo)
Active Pilots (5)
- Home Credit — Consumer lending (finalizing SOW)
- Globe — Telco/digital (2 pilots running)
- AdSpark — Ad agency (testing for clients)
- Inquiro — Research firm (co-development)
Qualified Pipeline
- AXA, Metrobank — Financial services
- Lenovo, HONOR — Consumer tech brands
- Riot Games — Gaming/esports
- DDB Group — Advertising network
Scientific Foundation
Academic Validation (Not "Science Theater")
arXiv Preprint 2505.22125v1 · Co-authored with University of the Philippines (Statistics & Psychology)
What we validated: Study compared simulated sentiment predictions vs. real survey responses from 2,485 nationally representative Filipinos across 12 consumer scenarios. Used NDAM methodology for representative sampling. Models built on HEXACO personality framework + Schwartz values (6 psychographic dimensions).
Key finding: Contextualized behavioral encoding achieved 88% quantitative agreement (QWA) with real human responses — significantly outperforming categorical methods (60% baseline, Wilcoxon p<0.0001). In live pilots, predictions matched real surveys within ±3 percentage points.
Why this matters for investors: Science creates customer trust (enterprises need validation to adopt AI) and demonstrates our approach works. But this is not our moat — moat is proprietary datasets + feedback loop + localization.
Market Opportunity
Beachhead → Expansion (Realistic TAM)
Year 1-2: Philippines
$600M
Serviceable, addressable
→
Year 3-4: SEA Regional
$3.2B
ID, VN, TH, MY
→
Long-term: Global
$147B
Total insights market
Beachhead (Now–2027): Philippine enterprise market research spend we can realistically capture: brands, agencies, research firms replacing surveys for pre-testing, concept validation, message optimization. We're targeting $600M serviceable market (enterprises spending >$10K USD/year on research). We operate in PHP locally but report in USD for fundraising.
Regional Expansion (2027-29): Indonesia, Vietnam, Thailand, Malaysia. Build localized datasets for each market. Combined serviceable market: $3.2B. This requires local partnerships + dataset development (18-24 months per market).
Long-term Vision: Global behavioral intelligence infrastructure layer. Comparable to Bloomberg (financial data) but for human behavior. Total global insights industry: $147B (ESOMAR). This is 10+ year horizon, not Series A pitch.
Next 12 Months
2026 Milestones (Conservative Targets)
Based on current pipeline + realistic conversion assumptions
Q1 2026 (LIVE)
Foundation
- 3 paying clients (₱70K MRR)
- 5 enterprise pilots running
- Globe/917V term sheet signed
- arXiv paper published
Q2 2026
Scale Sales
- Target: 8-10 total clients
- Convert 2 pilots to paid
- Hire 1 AE + 1 CS lead
- Launch image input (v2.0)
Q3 2026
Product + API
- Target: 15-18 total clients
- API beta for agencies
- Client data integration
- Indonesia research begins
Q4 2026
Series A Prep
- Target: 25+ total clients
- $1.8-2.7K MRR (₱100-150K PHP)
- Indonesia dataset complete
- Series A materials ready
Unit Economics
Actual Metrics from Current Contracts
Based on 3 paying clients + 5 pilots (early data, will evolve). Converting PHP to USD at ~₱55:$1 for investor clarity.
Average Contract Value
Current: $1.6K MRR / 3 clients (₱23K PHP avg). Range: $200-1.1K USD (₱10-60K PHP). Expect to trend toward $550-730/mo as we move upmarket.
Gross Margin
Compute: ~$110/client/mo at current scale. Should improve to 80%+ with volume. Customer support is main variable cost.
Customer Acquisition Cost
Blended across 3 closed deals (₱35K PHP). Founder-led sales + pilot conversions. Expect to rise to $910-1.1K with hired AEs but offset by higher ACV.
Payback Period
At current margin + ACV. Target: maintain <3 month payback as we scale. Early churn unknown (all clients <6 months old).
Team
Founding Team + Strategic Advisors
Axel Kornerup, MPA
CEO & Co-Founder
26+ years building companies. Founded 4 ventures: netopia (PH internet café chain, acquired), netVoice, netGames, netSolar. Part of founding team at InterVenn Biosciences (AI liquid biopsy, raised $100M+ from Catalio, DCVC, Khosla). Harvard Kennedy School MPA. Participated in >$500M enterprise value creation.
Jason Albia, MS
CSOO & Co-Founder
MS in AI systems. Previously led technical operations for enterprise SaaS at scale. Deep expertise in psychometric modeling, ML pipelines, scientific validation. Owns Predikta product architecture, research partnerships, and technical roadmap.
Advisory Board
Active Advisors
Aldo Carrascoso (InterVenn, >$1B outcomes) — fundraising + US expansion · Jojo Flores (Plug and Play, $10B+ portfolio) — enterprise BD + VC intros · Jane Walker (ex-PLDT/Singtel/San Miguel) — SEA partnerships + regional strategy
Research Partners
Academic Collaboration
University of the Philippines (Los Baños Statistics, Diliman Psychology) — co-authors on published research, ongoing collaboration on dataset expansion, validation studies, new behavioral frameworks. Provides academic credibility + talent pipeline.
Why Now
Timing & Market Forcing Functions
🤖
LLMs Enable Agentic Personas
Foundation models (GPT-4, Claude, Gemini) now sophisticated enough for nuanced behavioral simulation — impossible 3 years ago. We're riding the generative AI wave.
⚡
Speed Is Competitive Edge
Brands need weekly campaign iteration, not quarterly research cycles. Traditional agencies (Nielsen, Kantar) too slow. COVID accelerated this shift permanently.
💰
Budget Pressure
CMOs cutting research spend post-pandemic. 90% cost reduction vs. surveys is compelling ROI. Finance teams approve faster when savings are this stark.
🌏
SEA = Underserved + High Growth
Western tools expensive & not localized. 680M people in SEA but limited behavioral infrastructure. We own ground truth data for fastest-growing region.
🔬
Science = Trust Accelerator
Peer-reviewed validation creates enterprise trust. Electric Twin raised $14M without academic rigor — we have commercial traction + scientific credibility.
🚀
Category Validated by VCs
Electric Twin ($14M Atomico), Delphi, Synthetic Users — investors now understand behavioral AI category. No more education needed. We're commercial leader in SEA.
The Ask
Join us in building the behavioral intelligence layer for Southeast Asia
Raising $1.5M seed on a SAFE ($6M cap). We operate in PHP locally but raise in USD for investor alignment. Funds hire enterprise sales team (2 AEs + CS), expand to Indonesia/Vietnam (dataset development), and build self-serve product features. 18-month runway to Series A milestones: $1.5M ARR, 50+ customers, 3 markets live, clear path to profitability.