Inside VAKRA: Reasoning, Tool Use, and Failure Modes of Agents
•
28
AI & ML interests
Enterprise AI and ML, Foundation Models, Responsible AI
Recent Activity
Papers
IndustryAssetEQA: A Neurosymbolic Operational Intelligence System for Embodied Question Answering in Industrial Asset Maintenance
Efficient Agent Evaluation via Diversity-Guided User Simulation
Articles
pinned
Running
Agents
6
BlueBench Leaderboard
🥇
An open-source benchmark for enterprise use cases.
Running
1
ScarfBench
🐠
Java framework migration
Running
18
VAKRA Leaderboard
🏆
Evaluate AI agents on multi‑hop, multi‑source enterprise tasks
Running
2
BPO Benchmark Evaluation
📊
Run BPO recruiting benchmark with CUGA SDK
Sleeping
Agents
8
Ai Atlas Nexus Capabilities
📚
Intent to AI Capability with AI Atlas Nexus
Running
19
AssetOpsBench
🚀
Generate and benchmark machine learning models with ease
Running
7
ITBench-Lite-Space
🚀
Develop and run interactive code notebooks with JupyterLab
Running
Featured
100
CUGA Agent
🤖
Configurable Generalist Agent, leader in AppWorld Benchmark
Build error
Agents
2
3DGrid-VQGAN Demo
🐻
SMILES to 3D grid and reconstruction comparison demo
Running
4
AssetOpsBench
📉
Evaluating Autonomous AI Agents for Industry 4.0 Tasks
Build error
Agents
2
SMI-TED-demo1
🦦
Generate embeddings from SMILES strings or CSV files
Runtime error
Agents
2
Fm4m Eval Demo
👁
Runtime error
Agents
4
FM4M-demo2
🐢
Generate and analyze molecular structures
Running
Agents
7
FM4M-demo1
🐢
Generate and analyze molecular structures
Running
README
📈
Sleeping
Agents
4
Biomed.sm.mv Te 84m
👁
Prediction task tests for biomed-multi-view models