VEXKIN — INTELLIGENCE, EMBODIED.
Concept investor overview | 08 October 2026

THESIS
Useful humanoids need representations of the physical world, anticipation and a continuous learning loop. Vexkin's proposed platform connects a humanoid body, predictive intelligence and a simulation-to-evaluation workflow.

CONCEPT PRODUCT ARCHITECTURE
V1 Humanoid — A proposed platform for coordinated movement, vision-guided manipulation and human-supervised operation.
World Engine — A proposed intelligence layer for action-conditioned prediction, stage-aware planning and closed-loop replanning.
Vexkin Forge — A proposed workflow for demonstrations, scenario variation, evaluation and human feedback.

INITIAL APPLICATION DIRECTIONS
Manufacturing: component handling, machine tending and assembly support.
Logistics: sorting, tote transfers and controlled material movement.
Research: embodied learning, manipulation and human–robot interaction.

COMPANY-PROVIDED INFORMATION
Funding: USD 150,000 raised over the past 12 months, as of October 2026.
Academic backing named by Vexkin:
- Prof. Truyen Tran, Deakin University. Professor and Head of AI, Health and Science, Applied Artificial Intelligence Institute. Research themes include foundational AI, embodied intelligence, reasoning and planning.
  Profile: https://truyentran.github.io/
- Assoc. Prof. Quan Thanh Tho (Tho Quan), HCMUT, VNU-HCM. Dean of the Faculty of Computer Science and Engineering. Research interests include AI, NLP, intelligent systems and formal methods.
  Profile: https://www.cse.hcmut.edu.vn/qttho/doku.php?id=start
- Prof. Anthony Tung Kum Hoe, Department of Computer Science, School of Computing, National University of Singapore. Research interests include database systems, complex-data retrieval, data mining and collaborative analytics.
  Profile: https://www.comp.nus.edu.sg/cs/people/atung/
These institutional affiliations identify the individuals; they do not establish university sponsorship or a partnership with Vexkin. Formal Vexkin advisory roles are not specified.

PROPOSED NEXT MILESTONES
1. Validate a focused manipulation task with repeatable evaluation.
2. Connect predictive planning to a supervised robot control loop.
3. Scope an industrial pilot with measurable acceptance criteria.
Round size, valuation, use-of-funds allocation and commercial terms have not been specified.

RESEARCH REFERENCES
Figure, Helix 2.5 technical report, 17 September 2026:
https://www.figure.ai/news/helix-2-5-zero-shot-30-home-generalization
Liu et al., StageWAM, August 2026 preprint:
https://arxiv.org/abs/2608.10780
Mur-Labadia et al., V-JEPA 2.1, March 2026 preprint, revised June 2026:
https://arxiv.org/abs/2603.14482
Assran et al., V-JEPA 2, June 2025:
https://arxiv.org/abs/2506.09985

STATUS
This is a concept overview. Product names, renderings and the scripted planner demonstrate an intended direction, not released products or proven capabilities. No validated Vexkin specifications, performance benchmarks, customer deployments, patents, partnerships or revenues were supplied. Independent research outcomes are not Vexkin achievements. Funding and academic backing are company-reported.