Skip to content
Sign in
SWC 2026 FinalistGrand Final: Main Stage

Thetan

Dhaka research studio building AI “world models” — behavioral digital twins of places and machines — still at video-prototype stage.

Dhaka, Bangladesh, AsiaAI & software, World models and digital twinsBangladesh – regional final winner (Daffodil International University, announced 20 Sept. 2026)
Stage
UndeterminedAI at the core
Product
PrototypeNone known
Funding raised
Undisclosedno funding
Founded
20233 years ago
Founders
—
Global score
3.5Grade DRank 111 of 151

Overview

//thetan presents itself as a research-driven studio that builds “world models”: AI-powered digital twins that replicate real places and machines so ideas can be tested in simulation before touching the physical world. “Team Thetan” won the grand final of Startup World Cup Bangladesh 2026 at Daffodil International University (ahead of GreenPolaris Ltd and Shunno).

Traction

None known

No public evidence of traction found: only a digital twin video prototype (Dhanmondi, Dhaka) is shown.

Market

Digital twins

Digital twins are bought by asset-heavy operators (cities, utilities, plants, fleets) that want to test interventions in simulation before spending on physical change. Demand is accelerating because sensor data is cheap, GPU simulation is commoditising and generative “world model” research makes behavioral (not just geometric) twins feasible. The market is led by industrial software majors and platform ecosystems; newcomers win through a vertical niche and a proprietary dataset, usually via paid pilots. Thetan has not yet chosen a vertical or a paying customer.

Addressable market

2025
$21.1B
2030
$149.8B
+47.9% a year

Digital twins: $21.14bn in 2025 and $149.81bn in 2030 (CAGR 47.9%, MarketsandMarkets).

Customers

Urban planners and municipalities, infrastructure and utility operators, industrial and maritime equipment operators (target segments inferred; no customer documented)

Who signs the cheque

Engineering / digital-transformation heads at operators; public agencies for city-scale twins

Geographies

  • Bangladesh
  • South Asia
  • Global (software)

Drivers

  • Need to de-risk costly physical interventions through simulation
  • Falling cost of sensors, geodata and GPU compute
  • Rise of AI world models enabling behavior learning, not just 3D replicas

Headwinds

  • Incumbent platforms bundle twins with existing engineering software
  • Long, data-hungry enterprise pilots
  • No identified team, funding or customer to carry a sales cycle

MarketsandMarkets puts digital twins at $21.14bn in 2025 and $149.81bn in 2030 (CAGR 47.9%); Fortune $24.5bn, GMI $18.9bn.

Scalability 7/10. Replicable simulation software, but at research stage with no proven commercial product.

Market source: contentenginellc.com

Technology

Category

AI simulation software (digital twins / world models)

AI at the coreMachine learning / predictiveRobotics / physical AI

Maturity

  1. R&D
  2. Prototype
  3. Pilot
  4. On the market
  5. Scaling

prototype

How it works

A three-step pipeline ingests real-world data, learns the behavior of a place or machine, then simulates and predicts its response to changes. The only public artefact is a real-time twin demonstrator of the Dhanmondi neighborhood in Dhaka.

Moat

Weak today: the pipeline is generic and only shown as a video. Defensibility would come from proprietary behavioral datasets for one vertical and validated prediction accuracy.

Differentiation

Positions on behavior learning (how a system reacts) rather than static 3D/BIM replicas sold by most twin vendors; unproven.

Integration points

  • IoT / edge device fleets
  • Process-control historians and SCADA
  • GIS, satellite and geospatial layers
  • BIM / 3D scene formats

Depends on

  • Access to continuous real-world operating data
  • GPU compute and simulation tooling
  • Validation of predictive accuracy against ground truth

Use of AI

“World models”: AI-powered digital twins that learn the behavior of real places and machines in order to simulate them.

Product in detail

Three-step pipeline. real-world data ingestion, behavior learning, simulation/prediction; real-time digital twin demonstrator (video prototype “Dhanmondi | Prototype”, a Dhaka neighborhood).

Business

Model
Not disclosed; most plausible path is paid simulation studies and pilots, then per-twin software licences.
Revenue logic
Unknown — no pricing published.
Go-to-market
Regional-final visibility (SWC Bangladesh) to open pilots; would need a vertical design partner that owns operating data.

Next milestones

  1. Sign one paid vertical pilot with an operator that supplies live data
  2. Publish measured prediction accuracy versus actual outcomes
  3. Name the founding team and raise a first pre-seed ticket

Key risks

  • No team, funding or customer publicly identified
  • Generic positioning in a field crowded by large platforms
  • Data access dependent on partners

Opportunities

4 found in the network

Concrete moves for Thetan inside the network: who to sell to, sell through, build with or raise from, each with the reasoning, the evidence and the first step.

  • Co-development 3
  • Investors 1

Co-development3

An integration or joint product

Co-build a behavioral twin of shipboard LNG fuel systems

JJ&Companies already runs advanced process control on ships, i.e. it owns the operating data a behavioral twin needs. A Thetan twin trained on that data would let JJ test control strategies offline and sell predictive services to shipowners, while Thetan gets its first industrial dataset and reference.

30%confidence

Evidence from the records

  • “Three-step pipeline: real-world data ingestion, behavior learning, simulation/prediction”
  • “Advanced process control (J-IAS for LNG fuel gas supply systems)”
  • “J-IAS solution (LNG FGSS control) installed and commissioned on ships”
If it works
A paid proof-of-concept twin of one FGSS installation, giving JJ a simulation add-on and Thetan a first vertical reference.
First step
Send JJ&Companies a one-page proposal for a 6-week offline twin of one J-IAS installation using historical logs.

Add scenario simulation to building-level wildfire forecasts

Opal forecasts fire risk per building but does not describe what-if simulation; Thetan's behavior-learning engine could simulate spread and mitigation scenarios on top of Opal's risk layer. Opal gains a planning product for agencies and utilities; Thetan gains a funded application and geospatial data.

25%confidence

Evidence from the records

  • “proprietary Fire Potential Index (high-resolution hourly forecasts)”
  • “real-time digital twin demonstrator (video prototype “Dhanmondi | Prototype”, a Dhaka neighborhood)”
  • “FireVision product forecasts fire risk at building level up to 48 h in advance”
If it works
A joint scenario-simulation module demoed to one of Opal's agency or utility pilots.
First step
Book a technical call with Opal AI's CEO to test whether Thetan can ingest the Fire Potential Index as an input layer.

Use Qbee device fleets as the live data feed for twins

A twin is only as good as its data ingestion; Qbee already operates fleets of Linux edge devices with monitoring and remote access. An integration would let Thetan stream live device telemetry into its models and give Qbee customers a simulation layer on top of fleet management.

20%confidence

Evidence from the records

  • “real-world data ingestion”
  • “Unified control plane: configuration management, OTA updates, automation, monitoring and secure remote access to embedded devices.”
If it works
A Qbee-to-Thetan connector that shortens each pilot's data-integration phase.
First step
Open a free Qbee account (up to 5 devices) and prototype a telemetry export into the Thetan pipeline.

Investors1

A fund whose thesis fits

Pitch Iterative’s AI-focused pre-seed programme

Iterative writes $0.15–0.5M pre-seed/seed tickets for AI-focused founders from Southeast Asia “and beyond” and already backed network member Belli, which offers a warm intro. Thetan must first name its team and a pilot to be credible.

15%confidence

Evidence from the records

  • “Backs early-stage founders, increasingly AI-focused, from Southeast Asia and beyond.”
  • “Iterative (among others)”
If it works
A first institutional pre-seed ticket and accelerator support.
First step
Ask Belli's founders, via Champions Network, for an introduction to Iterative once a pilot LOI is signed.

Analyst reading of 21 Nov 2026, from the public record of the network. Method: analyst-ai.

Investors that fit

Singapore accelerator investing up to $500K in twice-yearly batches of increasingly AI-focused founders from Southeast Asia and beyond.

Bangladesh world models and digital twins — Asian AI deep tech, 'beyond SEA'.

40%confidence

Evidence from the records

  • “"AI & software"”
  • “World models and digital twins”
  • “No public evidence of traction found: only a digital twin video prototype (Dhanmondi, Dhaka) is shown.”
Intro path
Founder-to-founder intro via Belli (network portfolio company of Iterative).
First step
Apply with early-user evidence.

New Delhi angel network (450 members, 11 countries) with a ₹350 crore fund, backing Indian pre-seed and seed startups.

Bangladesh world models / digital twins — South Asian deep tech; IAN focus is India.

15%confidence

Evidence from the records

  • “"India"”
  • “World models and digital twins”
  • “No public evidence of traction found: only a digital twin video prototype (Dhanmondi, Dhaka) is shown.”
Intro path
Founder-to-founder intro via Astrome (network portfolio company of Indian Angel Network (IAN Group)).
First step
Only if an India entity is planned.

Indian fund that led Astrome's US$10M round (2025); thesis and ticket otherwise undocumented.

Bangladeshi world models and digital twins: South Asian deep tech, but only a prototype.

10%confidence

Evidence from the records

  • “World models and digital twins”
  • “Bangladesh”
Intro path
Astrome (Appollo portfolio)
First step
Not actionable until an India link and traction exist.

Turkey's national science and technology research council, which granted REPG Energy support through its BİGG programme.

Bangladeshi world-model studio at prototype stage: eligible only with a Turkish entity.

5%confidence

Evidence from the records

  • “World models and digital twins”
  • “Bangladesh”

Investment analysis

3.5D

Grade D

Rank 111 of 151

Disruption5.0/10
Business model1.0/10
Market6.0/10
Risk(lower is better)9.0/10

Global score = mean of disruption, model, market and (11 − risk). Last updated 21 Nov 2026.

Not investable.

Disruption

5/10

World models and digital twins: cutting-edge theme, no technical detail.

Business model

1/10

Only a video prototype.

Market

6/10

Score 6/10: TAM $21.14bn → base 6; scalability 7/10 (0); no commercial presence in ≥2 countries yet (0).

Risk

9/10

Lower is better

Very high risk: no information on the team or funding.

Strengths

  • Promising theme
  • Research-driven approach

Risks

  • No traction
  • Unknown team

Funding

Undisclosed, no funding

Amount used in the study. No public funding round found.

Note. no public funding round found

Regional win confirmed by The Daily Star (20/09/2026): Team Thetan was champion among 10 teams, with a total prize of 100,000 BDT shared among winners. The thetan.tech site names no founders, customers or funding; further searches were unsuccessful. Other fields kept.

Team

No founders listed yet.

Sources

Regional final
Bangladesh – regional final winner (Daffodil International University, announced 20 Sept. 2026)
Identification confidence
high
Founding year: 2023 (low confidence)
Site first archived in 2023, proxy (year the site existed, no confirmed founding date); content of the first snapshot not verified web.archive.org
  1. 1thedailystar.net/campus/noticeboard/news/grand-finale-startup-world-cup-bang
  2. 2newsbeep.com/us/855755/
  3. 3thetan.tech/
  4. 4youtube.com/watch?v=Go0fFSET5-o

From the open study Startup World Cup 2026 by TIBOK, based on public sources, verified in November 2026. Scores are an analyst's judgment on an explicit grid, not the jury's.

Wall

Wall

No posts yet