PeopleofSuperintelligence

Builder

Arnav Singhvi

Arnav Singhvi is a research engineer at Databricks whose work spans language-model programming, prompt optimization, and enterprise retrieval. He coauthored DSPy and LM Assertions, which give language-model pipelines programmable modules and constraints for self-refinement. In 2024, he presented DSPy Assertions while interning at Databricks and contributing to Stanford's NLP Group after graduating in computer science at UC Berkeley. His 2025 and 2026 work covers language-program benchmarking, medical vision-language prompt optimization, and grounded enterprise reasoning.

Works
13
Badges

Publication numbers from OpenAlex, as of 18 October 2026. Citations and h-index are hidden because this OpenAlex record may be incomplete or mix in other authors.

Timeline

  1. 2023

    DSPy and LM Assertions

    In 2023, he coauthors DSPy, a programming model for declarative language-model pipelines, and LM Assertions, a construct for adding constraints and self-refinement to those programs.

  2. March 2024

    A DSPy Assertions talk

    In March 2024, he presents a talk on DSPy Assertions while interning as a Research Scientist at Databricks. He is a recent computer-science graduate of UC Berkeley and a Stanford NLP Group researcher contributing to DSPy.

  3. 2025

    Benchmarking language programs

    At EMNLP 2025, he coauthors LangProBe, a large-scale benchmark spanning more than 2,000 combinations of language-program tasks, architectures, optimizers, and language models.

  4. 2025

    Prompt optimization in medicine

    In 2025, he coauthors Prompt Triage, which applies DSPy to automate prompt optimization for medical vision-language systems.

  5. March 2026

    A benchmark for enterprise agents

    In March 2026, he coauthors OfficeQA Pro, a benchmark for grounded AI-agent reasoning over U.S. Treasury Bulletins spanning nearly a century.

  6. Since 2025

    Research engineer at Databricks

    He is a research engineer at Databricks and works on Adaptive Instructed-Retriever, a model introduced in September 2026 that combines fast parallel retrieval with sequential search for complex, multi-hop questions.

Career

  1. Since 2025DatabricksResearch Scientist Intern, Research Engineer
  2. Until 2024Stanford UniversityResearcher

Education

  1. University of California, BerkeleyComputer Science degree

From the biography.

  1. The company share of AI researchers’ top papers rose from 3% to 32%Includes Databricks. We estimate where leading AI researchers worked when their most-cited papers appeared. The company share rose from 3% in 1975–1979 to 32% in 2020–2024.
  2. Stanford University trained or employed 27 of the 111 AI major-prize winners, more than any other universityIncludes Stanford University. Stanford University trained or employed 27 of the 111 AI major-prize winners, more than any other university. We compare doctorate and employment links for major-prize and best-paper winners.
  3. Google had the most people shaping AI every year from 2014 to 2025Includes Stanford University. Our estimates of which organizations employed the most people shaping AI from 1961 to 2025, with a year-by-year chart.
  4. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionIncludes Stanford University. Our estimates of where the people shaping AI studied and where their careers led, from education records in cited biographies.

Colleagues

People with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Arnav Singhvi yet. Send us a free-licensed photo.

  1. DSPy: Compiling Declarative Language Model Calls into Self-Improving Pipelines arxiv.org. Checked 18 October 2026.
  2. DSPy Assertions: Computational Constraints for Self-Refining Language Model Pipelines arxiv.org. Checked 18 October 2026.
  3. DSPy Assertions: Computational Constraints for Self-Refining Language Model Pipelines // Arnav Singhvi // AI in Production Talk home.mlops.community. Checked 18 October 2026.
  4. LangProBe: a Language Program Benchmark aclanthology.org. Checked 18 October 2026.
  5. Prompt Triage: Structured Optimization Enhances Vision-Language Model Performance on Medical Imaging Benchmarks arxiv.org. Checked 18 October 2026.
  6. OfficeQA Pro: An Enterprise Benchmark for End-to-End Grounded Reasoning arxiv.org. Checked 18 October 2026.
  7. Arnav Singhvi — Lacuna lacuna.tiptreesystems.com. Checked 18 October 2026.
  8. Adaptive Instructed-Retriever: Frontier-Quality Search at 2x Lower Latency databricks.com. Checked 18 October 2026.