Portrait of Alireza Salemi

Alireza Salemi

Research Scientist, Mosaic AI Research · Databricks

Google PhD Fellow · SIGIR Best Paper Awards (2023, 2024)

I am a Research Scientist at Databricks, working in Mosaic AI Research. My research sits at the intersection of information retrieval and large language models: making LLMs retrieve and use external knowledge reliably (retrieval-augmented generation), adapting them to the history and preferences of individual users (personalization), and composing them into agentic systems that search and reason over many steps. A common thread is grounding model behavior in evidence rather than parametric memory.

Before joining Databricks, I did my PhD in Computer Science at the University of Massachusetts Amherst, in the Center for Intelligent Information Retrieval, advised by Hamed Zamani and supported by a Google PhD Fellowship. I also spent time as a student researcher at Google DeepMind and Google Cloud AI. My work has appeared at ACL, EMNLP, NeurIPS, ICLR, SIGIR, WWW, and ICTIR, and received two SIGIR best paper awards.

Experience

Full experience
  1. Nov 2026 – Present

    Databricks

    Research Scientist, Mosaic AI Research

  2. Jun 2025 – Dec 2025

    Google Cloud AI

    Student Researcher

  3. Jun 2024 – May 2025

    Google DeepMind

    Student Researcher

Retrieval-Augmented Generation

Optimizing and evaluating the retrieval pipelines that ground LLMs in external knowledge.

eRAG DEDR uRAG Plan-and-Refine

LLM Personalization

Tailoring LLMs to individual users through benchmarks, reinforcement learning, self-training, and test-time scaling.

LaMP LaMP-QA RSPG ExPerT VAC PoT

Agentic AI Systems

Multi-agent and search-agent systems for autonomous information discovery and multi-step reasoning.

Blackboard GrepSeek Critic-R PR2 LiveRAG Marketplace

Long-context and Memory

Benchmarking and improving LLM memory over millions of tokens of personal context.

BEAM

Recent news

All news
  1. Nov 2026
    Joined Databricks as a Research Scientist in Mosaic AI Research
  2. Jun 2026
    Won 2nd place at the Grounded Reasoning Cup (Data + AI Summit 2026), earning $30K in Databricks AI credits
  3. 2026
    3 papers accepted at SIGIR 2026
  4. 2026
    Paper accepted at ICLR 2026: Beyond a Million Tokens: Benchmarking and Enhancing Long-Term Memory in LLMs
  5. 2026
    Paper accepted at WWW 2026: Pathways of Thoughts: Multi-Directional Thinking for Long-form Personalized Question Answering

Selected publications

All publications
  1. GrepSeek: Training Search Agents for Direct Corpus Interaction

    Alireza Salemi, Chang Zeng, Atharva Nijasure, Jui-Hui Chung, Razieh Rahimi, Fernando Diaz, Hamed Zamani

    arXiv

  2. LLM-based Multi-Agent Blackboard System for Information Discovery in Data Science

    Alireza Salemi, Mihir Parmar, Palash Goyal, Yiwen Song, Jinsung Yoon, Hamed Zamani, Hamid Palangi, Tomas Pfister

    NeurIPS 2026

  3. LaMP: When Large Language Models Meet Personalization

    Alireza Salemi, Sheshera Mysore, Michael Bendersky, Hamed Zamani

    ACL 2024

  4. Evaluating Retrieval Quality in Retrieval-Augmented Generation

    Alireza Salemi, Hamed Zamani

    SIGIR 2024Best Short Paper Award

  5. Optimization Methods for Personalizing Large Language Models through Retrieval Augmentation

    Alireza Salemi, Surya Kallumadi, Hamed Zamani

    SIGIR 2024

  6. A Symmetric Dual Encoding Dense Retrieval Framework for Knowledge-Intensive Visual Question Answering

    Alireza Salemi, Juan Altmayer Pizzorno, Hamed Zamani

    SIGIR 2023Best Student Paper Award