Evolve.Pub
Sign inScan your paper
MG

Manish Gupta

Microsoft (United States)

Unclaimed · AI-structured profileOpenAlex

Microsoft NLP researcher: LLM agents, agentic web search, and benchmarks for evaluation and text classification

4

papers here

2

reviewed by Evolve

1

first-author

0

corresponding

—

works · OpenAlex

—

h-index · OpenAlex

Bibliometric figures come from OpenAlex (2026-10-17) and cover the author's entire public record, not just the papers listed here. Reproducibility signal comes from Evolve's own reviews.

Research profile

Output: insufficient dataOutputImpact: insufficient dataImpactFocus: 0/100FocusCollaboration: 61/100CollaborationActivity: insufficient dataActivityReproducibility: 38/100Reproducibility

Output—

No output metrics are available in this profile: publication counts and venue-level detail are absent from the snapshot, so this axis is unscored. It would measure the volume, consistency, and venue quality of the researcher's scholarly output over time.

Impact—

Impact cannot be assessed here: no citation counts, h-index, or venue-influence data are provided for this researcher in the current snapshot. This axis would measure how widely the work is cited, adopted, and built upon by other researchers; it remains unscored.

Focus0/100

The radar focus score is 0, consistent with a field distribution split exactly 50/50 between 'NLP / information retrieval (LLM agents)' and 'natural language processing (LLM evaluation and benchmarking)'. The four catalog papers also spread across agentic web search measurement, travel-planning benchmarks, text classification, and multi-agent planning — breadth across adjacent LLM topics rather than one narrow line.

Collaboration61/100

Collaboration scores 61 on the radar, built on 20 distinct coauthors, with a recurring circle of twelve named collaborators including Priyanshu Karmakar, Shubhojit Mallick, Abhik Jana, and Shreya Ghosh. Catalog papers place him first, middle, and last in the author list, none as corresponding author, indicating rotating team-based projects rather than one fixed lab group.

Activity—

Activity data is insufficient: the snapshot contains no year-by-year publication series beyond the four 2026 catalog papers, so publication cadence and recent-productivity trends cannot be computed. This axis would measure output frequency and continuity of research engagement over time; it is left unscored.

Reproducibility38/100

Reproducibility scores 38, computed as a mean of 38 across 2 scored papers. Within the catalog, Evolve rates reproducibility readiness low for the agentic web search study and medium for both UTP-Bench and the confusion-aware classification paper, suggesting benchmark- and prompt-based methods whose artifacts and evaluation protocols are only partly standardized.

Research summary

Manish Gupta, listed with Microsoft (United States), is a natural language processing researcher whose Evolve catalog spans four 2026 arXiv papers at the intersection of LLM agents, information retrieval, and evaluation methodology. His field distribution splits evenly (50/50) between NLP/information retrieval on LLM agents and NLP on LLM evaluation and benchmarking. Three strands recur. First, agentic web search: 'Characterizing Web Search by Conversational LLM Agents' is described by Evolve as the first cross-platform measurement of agentic web search, covering four platforms. Second, uncertainty-aware travel planning: UTP-Bench injects empirical transit-delay and crowd-density uncertainty into LLM itinerary assessment. Third, retrieval-augmented classification: the first-authored 'From Confusion to Clarity' pairs error-driven confusion partners with LLM-generated disambiguation rules to raise LLM classification F1. A fourth paper, TRIPPULSE, applies multi-agent travel planning with review-grounded reasoning. Radar scores place collaboration at 61 (20 distinct coauthors), reproducibility at 38 (mean over two scored papers), and focus at 0, matching the evenly split field distribution. Evolve's verdicts flag open questions: judge-model heterogeneity in the search study, annotation-metric circularity in UTP-Bench, and uneven, costly gains in the classification work.

Collaboration pattern

Gupta's catalog work involves 20 distinct coauthors, with twelve recurring collaborators including Priyanshu Karmakar, Shubhojit Mallick, Abhik Jana, Chaitanya Giri, Jayasimha Talur, Borru Vijay Sai, Shreya Ghosh, Mahsa Amani, Seungeon Lee, Abhisek Dash, Asmaa El Fraihi, and Yunah Jang. Author positions rotate — first on the confusion-aware classification paper, middle on two benchmark studies, last on TRIPPULSE — and no catalog paper lists him as corresponding author, suggesting project-based teamwork rather than a single fixed lab group.

Research keywords

LLM agentsAgentic web searchLLM benchmarkingTravel planningUncertainty-aware evaluationItinerary generationText classificationRetrieval augmentationKnowledge injectionConfusion matrixPrompt engineeringMulti-agent systems

Highlights

Timeline

  1. 2026Publishes first-authored 'From Confusion to Clarity', combining confusion-aware retrieval with LLM-generated knowledge injection for text classification.
  2. 2026Releases UTP-Bench, an uncertainty-aware travel planning benchmark incorporating transit delay and crowd density factors.
  3. 2026Co-authors 'Characterizing Web Search by Conversational LLM Agents', positioned as the first cross-platform study of agentic web search across four platforms.
  4. 2026Appears as last author on TRIPPULSE, a multi-agent travel planning system with review-grounded reasoning (not yet reviewed by Evolve).

Papers in the Evolve catalog

Frequent co-authors

Priyanshu Karmakar×2Shubhojit Mallick×2Abhik Jana×2Chaitanya Giri×1Jayasimha Talur×1Borru Vijay Sai×1Shreya Ghosh×1Mahsa Amani×1Seungeon Lee×1Abhisek Dash×1Asmaa El Fraihi×1Yunah Jang×1

Is this you? Claim this profile

Enter the email you published with. We'll send a one-time claim link; after claiming you can edit the introduction or hide this page. Requests from a mismatched address go to manual review.

About this profile

This profile is AI-structured by Evolve.Pub from public bibliographic data (OpenAlex) and Evolve's own paper reviews. It is not written or endorsed by the researcher unless the profile is claimed. Radar dimensions are computed from real bibliometrics with fixed formulas — raw figures are shown alongside; missing dimensions mean insufficient data, never zero. Errors? Claim the profile to correct or hide it.

  • Bibliometric radar scores are computed from the platform's public bibliometric snapshot only; no citation or venue-level source data is included in this profile.
  • Output, impact, and activity axes are unscored because the underlying metrics (publication counts, citations, yearly productivity series) are not available in the current snapshot.
  • 'Manish Gupta' is a common name; this profile is name-resolved from public bibliographic data and institution affiliation, and may not represent a single individual.
  • All four catalog papers are 2026 arXiv preprints; TRIPPULSE has no Evolve review yet, and the reproducibility score covers only 2 of the papers.
Manish Gupta — Researcher Profile | Evolve.Pub