Centific · Applied AI Research
Frontier AI research,
grounded in the real world.
Centific Applied AI Research builds the benchmarks, RL environments and expert data that show how frontier models behave outside the lab — and how to make them better.
32
Papers
16
Leaderboards
7
Verticals
8
NeurIPS ’26 papers
Applied AI Research
Leaderboards
RL Environments
Publications
Research Insights
Jobs
NeurIPS 2026 · 8 accepted papers →
02
Leaderboards
16 benchmarks · 7 verticals
Benchmarks for the domains where frontier models still fall short
Open leaderboards built with domain experts across healthcare, audio, safety, physical AI and agentic RL.
#
Benchmark
Vertical
Scope
01
Med-ART
Clinical agent evaluation on a FHIR-compliant virtual EHR
Healthcare
9 models · 1,200 tasks
→
02
MedMosaic
Medical audio reasoning across cough, heart, lung and speech
Audio
13 models · 11 metrics
→
03
FLINT
Multi-turn adversarial safety red-teaming
Safety
11 models · 1,406 conversations
→
04
B4 Dexterous Assembly
Teleoperation vs. simulation for robotic dexterity on the Unitree G1
Physical AI
29 datasets · 1,413 episodes
→
05
BA Agent Bench
Enterprise requirements and user-story generation
Agentic & RL
8 models · 119 user stories
→
03
Publications
32 papers
Peer-reviewed research from the lab
Papers on evaluation, agents, safety, speech and physical AI, with the datasets and benchmarks behind them.
Date
Title
Verticals
Authors
Oct 02, 2026
Same Output, Different Gold: Measuring How Reference Choice Moves a Multilingual Benchmark Score
Internationalization
Parth Kulshreshtha, Shivali Dalmia, Abhishek Mukherji
Oct 02, 2026
CuBEs: Culturally-Situated Behavioral Evaluations and the Limitations of Culture-Blind LLM Judges
Safety
Internationalization
Hoda Ayad, Tanu Mitra, Abhishek Mukherji
Sep 25, 2026
Closing the Deal Is Not Enough: A Behavioral Rubric for Agent-to-Agent Marketplaces
Agentic & RL
Shivali Dalmia, Ashi Jain, Abhishek Mukherji
Sep 22, 2026
Beyond End-Task Success: How to Audit Visual Experience Retrieval in Robotics
Physical AI
Eshika Pathak, Leela Krishna
Sep 18, 2026
SEABED: SouthEast Asian Benchmark for Evaluating Audio Reasoning
Audio
Internationalization
Harshit Rajgarhia, Asif Shaik, Rachuri Lokesh, Sushanta Kumar Pani, Abhishek Mukherji
Sep 18, 2026
When Should a Failing Robot Ask? Initiating Corrective Human-Robot Dialogue from Audited Sensor Evidence
Physical AI
Eshika Pathak, Leela Krishna
04
Research insights
From the blog
What we tested, what we found, and why it matters
Plain-language write-ups of our papers and experiments for teams evaluating and shipping AI.
05
Join the lab
6 open roles
Work on the evaluation, RL and data problems frontier labs care about
We’re hiring research scientists, engineers and interns across reinforcement learning, LLM evaluation and speech.
Senior Staff Research Scientist, Reinforcement Learning
Reinforcement learning
Remote (USA)
↗
Staff Research Scientist, Reinforcement Learning
Reinforcement learning
Remote (USA)
↗
Research Scientist, LLM Evaluation & Post-Training
Evaluation & post-training
Remote (USA)
↗
Applied Reinforcement Learning Engineer
Reinforcement learning
Remote (USA)
↗
Connect with Centific
Updates from the frontier of AI data.
Receive updates on platform improvements, new workflows, evaluation capabilities, data quality enhancements, and best practices for enterprise AI teams.