David Vázquez

Bio

Tue, 07 Apr 2026 00:00:00 +0000

Contact

Tue, 07 Apr 2026 00:00:00 +0000

Experience

Tue, 07 Apr 2026 00:00:00 +0000

Projects

Tue, 07 Apr 2026 00:00:00 +0000

Publications

Tue, 07 Apr 2026 00:00:00 +0000

We're hiring researchers and engineers

Tue, 07 Apr 2026 00:00:00 +0000

The Foundational AI Research team at ServiceNow Research is hiring Senior Research Engineers and Scientists in Montreal. Research areas include AI agents, LLMs, reinforcement learning, time series analysis, and security. Apply here.

Presenting at ICLR 2026 in Rio de Janeiro

Wed, 01 Apr 2026 00:00:00 +0000

Our team presented multiple papers at ICLR 2026 in Rio de Janeiro, including work on multimodal models, web agents, and enterprise AI benchmarks.

Apriel-1.5-OpenReasoner: RL Post-Training for General-Purpose and Efficient Reasoning

Thu, 01 Jan 2026 00:00:00 +0000

StarFlow: Generating Structured Workflow Outputs from Sketch Images

Thu, 01 Jan 2026 00:00:00 +0000

VectorGym: A Multitask Benchmark for SVG Code Generation, Sketching, and Editing

Thu, 01 Jan 2026 00:00:00 +0000

WildSVG: Towards Reliable SVG Generation Under Real-Word Conditions

Thu, 01 Jan 2026 00:00:00 +0000

AI Tools for Indigenous Languages

Sat, 01 Nov 2025 00:00:00 +0000

An NSERC Discovery Grant funded project developing multimodal AI tools for underrepresented languages, with a focus on the Matsigenka language of Peru and Inuktitut in northern Canada. The project follows OCAP and TCPS 2 data sovereignty principles and involves community partners including Tejiendo Puentes en Salud, Ayni Desarrollo, Heritage Lab, and CECONAMA. Conducted through Polytechnique Montréal in collaboration with MILA.

NSERC Discovery Grant awarded for Indigenous language AI tools

Sat, 01 Nov 2025 00:00:00 +0000

Awarded an NSERC Discovery Grant funding research on multimodal AI translation and literacy tools for the Matsigenka and Inuktitut communities, conducted through Polytechnique Montréal in collaboration with MILA.

AlignVLM accepted at NeurIPS 2025

Mon, 01 Sep 2025 00:00:00 +0000

Our paper “AlignVLM: Bridging Vision and Language Latent Spaces for Multimodal Understanding” has been accepted at NeurIPS 2025.

Appointed Adjunct Professor at Polytechnique Montréal

Mon, 01 Sep 2025 00:00:00 +0000

Joined Polytechnique Montréal as Adjunct Professor affiliated with MILA, co supervising doctoral students and leading research on multimodal AI tools for Indigenous language preservation.

EnterpriseOps-Gym

Tue, 01 Apr 2025 00:00:00 +0000

EnterpriseOps-Gym features 1,150 expert designed tasks across 8 interconnected enterprise domains, with persistent state, strict verification logic, and policy aware execution requirements. It tests whether AI agents can handle domain expertise, not just general reasoning.

EnterpriseOps-Gym Released

Tue, 01 Apr 2025 00:00:00 +0000

Our new benchmark for evaluating stateful agentic planning in enterprise settings is now public. 1,150 tasks across 8 domains. Paper · Code

Apriel Model Family

Sat, 01 Mar 2025 00:00:00 +0000

The Apriel family of open language models developed at ServiceNow Research, including base models (Apriel 1.5, 1.6), reasoning models (AprielReasoner), and safety models (AprielGuard, an 8B parameter guardian model).

StarVector accepted at CVPR 2025

Sat, 01 Mar 2025 00:00:00 +0000

Our paper “StarVector: Generating Scalable Vector Graphics Code from Images” has been accepted at CVPR 2025. Former intern Juan A. Rodriguez co-founded QuiverAI based on this research, raising an $8.3M seed round led by Andreessen Horowitz.

BigDocs accepted at ICLR 2025

Wed, 22 Jan 2025 00:00:00 +0000

Our paper “BigDocs: An Open Dataset for Training Multimodal Models on Document and Code Tasks” has been accepted at ICLR 2025. Paper · Code

AgentAda: Skill-Adaptive Data Analytics for Tailored Insight Discovery

Wed, 01 Jan 2025 00:00:00 +0000

AlignVLM: Bridging Vision and Language Latent Spaces for Multimodal Document Understanding

Wed, 01 Jan 2025 00:00:00 +0000

BigCharts-R1: Enhanced Chart Reasoning with Visual Reinforcement Finetuning

Wed, 01 Jan 2025 00:00:00 +0000

BigDocs

Wed, 01 Jan 2025 00:00:00 +0000

BigDocs is a large scale, open, and permissively licensed dataset for training multimodal models on document understanding and code generation tasks. Published at ICLR 2025.

BigDocs: An Open Dataset for Training Multimodal Models on Document and Code Tasks

Wed, 01 Jan 2025 00:00:00 +0000

Distilling Specialized Orders for Visual Generation

Wed, 01 Jan 2025 00:00:00 +0000

Grounding Computer Use Agents on Human Demonstrations

Wed, 01 Jan 2025 00:00:00 +0000

Intent Discovery using Large Language Models

Wed, 01 Jan 2025 00:00:00 +0000

Rendering-Aware Reinforcement Learning for Vector Graphics Generation

Wed, 01 Jan 2025 00:00:00 +0000

StarVector: Generating Scalable Vector Graphics Code from Images and Text

Wed, 01 Jan 2025 00:00:00 +0000

UI-Vision: A Desktop-Centric GUI Benchmark for Visual Perception and Interaction

Wed, 01 Jan 2025 00:00:00 +0000

WebMMU: A Benchmark for Multimodal Multilingual Website Understanding and Code Generation

Wed, 01 Jan 2025 00:00:00 +0000

Improved Training Set Selection for Semi-Supervised Learning

Sun, 01 Dec 2024 00:00:00 +0000

WorkArena accepted at ICML 2024

Mon, 08 Jul 2024 00:00:00 +0000

Our paper “WorkArena: How Capable are Web Agents at Solving Common Knowledge Work Tasks?” has been accepted at ICML 2024. Paper · Code

WorkArena and BrowserGym

Mon, 01 Jul 2024 00:00:00 +0000

WorkArena is a benchmark of tasks based on the ServiceNow platform that measures how well web agents can perform common knowledge work. BrowserGym provides a rich environment for designing and evaluating such agents with multimodal observations and a comprehensive action set. Published at ICML 2024.

A Multimodal Class-Incremental Learning Benchmark for Classification Tasks

Mon, 01 Jan 2024 00:00:00 +0000

CADet: Fully Self-Supervised Out-of-Distribution Detection with Contrastive Learning

Mon, 01 Jan 2024 00:00:00 +0000

GEO-Bench: Toward Foundation Models for Earth Monitoring

Mon, 01 Jan 2024 00:00:00 +0000

Group Robust Classification without Any Group Information

Mon, 01 Jan 2024 00:00:00 +0000

InsightBench: Evaluating Business Analytics Agents through Multi-Step Insight Generation

Mon, 01 Jan 2024 00:00:00 +0000

RepLiQA: A Question-Answering Dataset for Benchmarking LLMs on Unseen Reference Content

Mon, 01 Jan 2024 00:00:00 +0000

Towards Good Validation Metrics for Generative Models in Offline Model-Based Optimisation

Mon, 01 Jan 2024 00:00:00 +0000

WorkArena: How Capable Are Web Agents at Solving Common Knowledge Work Tasks?

Mon, 01 Jan 2024 00:00:00 +0000

XC-Cache: Cross-Attending to Cached Context for Efficient LLM Inference

Mon, 01 Jan 2024 00:00:00 +0000

Anomaly Detection using Graph Neural Networks