PDF Viewer

BUILDER'S SANDBOX

Build This Paper

Use an AI coding agent to implement this research.

OpenAI Codex
OpenAI CodexAI Agent

Lightweight coding agent in your terminal.

Claude Code
Claude CodeAI Agent

Agentic coding tool for terminal workflows.

AntiGravity IDE
AntiGravity IDEScaffolding

AI agent mindset installer and workflow scaffolder.

Cursor
CursorIDE

AI-first code editor built on VS Code.

VS Code
VS CodeIDE

Free, open-source editor by Microsoft.

MVP Investment

$9K - $12K
6-10 weeks
Engineering
$8,000
Cloud Hosting
$240
SaaS Stack
$300
Domain & Legal
$100

6mo ROI

2-4x

3yr ROI

10-20x

Lightweight AI tools can reach profitability quickly. At $500/mo average contract, 20 customers = $10K MRR by 6mo, 200+ by 3yr.

Talent Scout

I

Iman Ahmadi

Sharif University of Technology

M

Mehrshad Taji

Sharif University of Technology

A

Arad Mahdinezhad Kashani

Sharif University of Technology

A

AmirHossein Jadidi

Sharif University of Technology

Find Similar Experts

Robotic experts on LinkedIn & GitHub

References (46)

[1]
LERa: Replanning with Visual Feedback in Instruction Following
2025Svyatoslav Pchelintsev, Maxim A. Patratskiy et al.
[2]
Vision-Language-Action (VLA) Models: Concepts, Progress, Applications and Challenges
2025Ranjan Sapkota, Yang Cao et al.
[3]
DaDu-E: Rethinking the Role of Large Language Model in Robotic Computing Pipeline
2024Wenhao Sun, Sai Hou et al.
[4]
MALMM: Multi-Agent Large Language Models for Zero-Shot Robotic Manipulation
2024Harsh Singh, Rocktim Jyoti Das et al.
[5]
Synergistic Simulations: Multi-Agent Problem Solving with Large Language Models
2024Asher Sprigler, Alexander Drobek et al.
[6]
ReKep: Spatio-Temporal Reasoning of Relational Keypoint Constraints for Robotic Manipulation
2024Wenlong Huang, Chen Wang et al.
[7]
ReplanVLM: Replanning Robotic Tasks With Visual Language Models
2024Aoran Mei, Guo-Niu Zhu et al.
[8]
Wonderful Team: Zero-Shot Physical Task Planning with Visual LLMs
2024Zidan Wang, Rui Shen et al.
[9]
Robots Can Multitask Too: Integrating a Memory Architecture and LLMs for Enhanced Cross-Task Robot Action Generation
2024Hassan Ali, Philipp Allgeuer et al.
[10]
BadRobot: Jailbreaking Embodied LLMs in the Physical World
2024Hangtao Zhang, Chenyu Zhu et al.
[11]
Robotic Control via Embodied Chain-of-Thought Reasoning
2024Michał Zawalski, William Chen et al.
[12]
Generate Subgoal Images Before Act: Unlocking the Chain-of-Thought Reasoning in Diffusion Model for Robot Manipulation with Multimodal Prompts
2024Fei Ni, Jianye Hao et al.
[13]
OpenVLA: An Open-Source Vision-Language-Action Model
2024Moo Jin Kim, Karl Pertsch et al.
[14]
RoboMamba: Efficient Vision-Language-Action Model for Robotic Reasoning and Manipulation
2024Jiaming Liu, Mengzhen Liu et al.
[15]
MOKA: Open-World Robotic Manipulation through Mark-Based Visual Prompting
2024Kuan Fang, Fangchen Liu et al.
[16]
PIVOT: Iterative Visual Prompting Elicits Actionable Knowledge for VLMs
2024Soroush Nasiriany, Fei Xia et al.
[17]
Large Language Models Are Neurosymbolic Reasoners
2024Meng Fang, Shilong Deng et al.
[18]
Large Language Models for Robotics: Opportunities, Challenges, and Perspectives
2024Jiaqi Wang, Zihao Wu et al.
[19]
RePLan: Robotic Replanning with Perception and Language Models
2024Marta Skreta, Zihan Zhou et al.
[20]
PERIA: Perceive, Reason, Imagine, Act via Holistic Language and Vision Planning for Manipulation
2024Fei Ni, Jianye Hao et al.

Showing 20 of 46 references

Founder's Pitch

"MALLVi offers a multi-agent robotic manipulation framework integrating language and vision models for adaptive task execution in dynamic environments."

Robotic ManipulationScore: 7View PDF ↗

Commercial Viability Breakdown

0-10 scale

High Potential

1/4 signals

2.5

Quick Build

4/4 signals

10

Series A Potential

2/4 signals

5

Sources used for this analysis

arXiv Paper

Full-text PDF analysis of the research paper

GitHub Repository

Code availability, stars, and contributor activity

Citation Network

Semantic Scholar citations and co-citation patterns

Community Predictions

Crowd-sourced unicorn probability assessments

Analysis model: GPT-4o · Last scored: 2/18/2026

🔭 Research Neighborhood

Generating constellation...

~3-8 seconds

Why It Matters

MALLVi matters because it addresses the limitations of existing robotic manipulation systems by providing a feedback-driven, multi-agent framework that improves reliability and adaptability in dynamic environments.

Product Angle

This framework can be productized as a software solution for robotics companies looking to enhance their robot's task execution capabilities with advanced feedback mechanisms, reducing failure rates in unstructured environments.

Disruption

MALLVi could replace existing robotic systems that operate primarily in open-loop without effective real-time feedback, offering more adaptable and reliable solutions.

Product Opportunity

The market for automation and robotics in logistics and manufacturing is vast, where companies are looking to improve efficiency and reliability of task execution under varying conditions.

Use Case Idea

A commercial application could be automated warehouse robots that adapt to dynamically changing environments for tasks like sorting and handling diverse items following human-like instructions.

Science

MALLVi leverages a multi-agent approach where different specialized agents handle distinct tasks in robotic manipulation, such as task decomposition, scene understanding, and error correction, using language and vision models for feedback and improvement.

Method & Eval

MALLVi was tested in simulated environments (VIMABench, RLBench) and real-world settings, showing improved success rates in zero-shot manipulation tasks compared to prior methods.

Caveats

Challenges include potential integration issues with existing robotic systems, especially if they rely on specific proprietary technologies, and the need for robust testing in diverse real-world scenarios.

Author Intelligence

Iman Ahmadi

LEAD
Sharif University of Technology
iman.ahmadi@ee.sharif.edu

Mehrshad Taji

Sharif University of Technology
mehrshad.taji@ee.sharif.edu

Arad Mahdinezhad Kashani

Sharif University of Technology
arad.mnk81@sharif.edu

AmirHossein Jadidi

Sharif University of Technology
jadidi@ee.sharif.edu

Saina Kashani

Sharif University of Technology
saina kashani@ee.sharif.edu

Babak Khalaj

Sharif University of Technology
khalaj@ee.sharif.edu