Adeep Hande
Adeep Hande

Adeep Hande

ML Research Engineer, Comcast Applied AI Research · Washington, D.C.

Research

How do we make large-scale personalization both efficient and reliable?

I work on LLM post-training and generative recommendation, with a growing focus on making LLM inference fast and cheap enough to run at real production scale. At Comcast, this means building knowledge distillation and post-training pipelines for user profile generation, using LLMs to solve cold-start recommendation, and serving all of it efficiently enough to run over 5 to 10M users.

Generative & sequential recommendation

LLM-generated synthetic interactions and LLM-guided negative selection for cold-start and long-tail items (CIKM 2026, EACL 2027, NAACL 2027).

LLM post-training at scale

SFT, DPO, and distillation to train small models from larger teachers, cheap enough to deploy (NAACL 2027).

Efficient inference

Self-hosted vLLM serving that runs LLM personalization at roughly 650× lower cost than managed APIs.

News

Selected Papers

UNDER REVIEW
From Head to Tail: LLM-Augmented Sequential Recommendation and Trade-off Optimization
Adeep Hande, Shaghayegh Agah, Neeraj Sharma, Yejin Kim, Karina H. Halevy, Sardar Hamidian
EACL 2027 Under review PDF
LLM-Driven Segment Discovery: Uncovering Latent User Intent at Scale
Shaghayegh Agah*, Adeep Hande*, Neeraj Sharma (* equal contribution)
NAACL 2027, Industry Track Under review PDF
Candidate Grounding Matters: LLM-Guided Negative Selection for Production Recommenders
Yejin Kim, Adeep Hande, Shaghayegh Agah, Neeraj Sharma, Sardar Hamidian
NAACL 2027, Industry Track Under review PDF
PUBLISHED
From Head to Tail: LLM-Augmented Sequential Recommendation and Trade-off Optimization
Adeep Hande, Shaghayegh Agah, Neeraj Sharma, Karina Halevy, Sardar Hamidian
GRAIL Workshop @ CIKM 2026 PDF
Evaluating ASR Quality at Scale on TV Entertainment Platforms
Adeep Hande, Kishorekumar Sundararajan, Yidnekachew Endale, Akshatha Bapu KrishnaSwamy, Sachin Dabral, Dawn Reed, Michael Pereira
GEM Workshop @ ACL 2026 PDF
Simulating Discoverability for Upcoming Content on Entertainment Platforms
RecSys 2025, Prague Paper
Content Moderation in TV Search: Balancing Compliance, Relevance, and User Experience
SIGIR 2025 Paper
Show all papers (26) →

Earlier Work

From September 2020 to January 2022, I worked with Prof. Bharathi Raja Chakravarthi on NLP for low-resource Dravidian languages.

Multi-task Learning in Under-resourced Dravidian Languages
Adeep Hande, Siddhanth U. Hegde, Bharathi Raja Chakravarthi
Journal of Data, Information and Management (Springer), 2022 Paper
The Best of Both Worlds: Dual Channel Language Modeling for Hope Speech Detection in Low-resourced Kannada
Adeep Hande, Siddhanth U. Hegde, Sivanesan Sangeetha, Ruba Priyadharshini, Bharathi Raja Chakravarthi
LT-EDI Workshop @ ACL 2022
KanCMD: Kannada CodeMixed Dataset for Sentiment Analysis and Offensive Language Detection
Adeep Hande, Ruba Priyadharshini, Bharathi Raja Chakravarthi
PEOPLES Workshop @ COLING 2020
All earlier work →

Experience

ML Research Engineer, Comcast

Recommendation and Arbitration team · Washington, D.C.
  • LLM post-training for user profile generation: built a teacher-student post-training pipeline for a 3B-parameter student model, using SFT with auxiliary reasoning objectives, DPO for alignment, and off- and on-policy evaluation. The fine-tuned models then act as teachers for downstream recommendation tasks.
  • LLM data augmentation for cold-start recommendation: designed an LLM-based pipeline injecting synthetic user histories for cold-start items, achieving an 80% relative NDCG improvement while maintaining overall quality (CIKM 2026, EACL 2027).
  • Efficient inference at scale: built a self-hosted vLLM serving setup on 4×A100 GPUs that reaches about 326k tokens/sec for user profile inference across 5 to 10 million users, roughly 650× cheaper than managed APIs at that scale.
  • Xfinity Campaign Manager forecasting: built campaign-level time-series forecasting models and ETL pipelines (Spark, Databricks, SQL) across EU and NA markets, with automated dashboarding for weekly media-buy pacing and budget allocation.
  • Content discoverability simulator: architected a large-scale simulation using LLM-generated synthetic queries and historical clustering to evaluate upcoming content in the Xfinity TV ecosystem (RecSys 2025).

Machine Learning Graduate Research Intern, Comcast

Washington, D.C.
  • Xfinity ASR evaluation tool: built an end-to-end speech evaluation framework benchmarking ASR quality (WER, embedding-based metrics) across diverse user demographics and acoustic conditions, scaling annotation throughput from 1K/day to 100K/week via automated proxy metrics.
  • Search quality evaluation: deployed a hybrid system combining heuristics and LLMs for real-time moderation of 5M+ daily voice queries; introduced a time-aware dynamic lexicon reducing false positives by 35% (SIGIR 2025).

Education

Indiana University Bloomington

M.S. in Data Science

Indian Institute of Information Technology, Tiruchirappalli

B.S. in Computer Science

Skills

ML & NLPLarge language models, LLM fine-tuning (SFT, DPO, RLHF), knowledge distillation, recommendation systems, ranking, RAG, information retrieval, content moderation, bias & fairness, ASR evaluation
FrameworksPyTorch, TensorFlow, Hugging Face Transformers, vLLM, LangChain, OpenAI API, MLflow, Airflow
Data & infraSpark, Databricks, AWS (SageMaker, Lambda, EC2), Docker, FastAPI, MongoDB, Kafka
LanguagesPython, SQL

Academic Service

100+ peer reviews across academic venues including InterSpeech, SIGIR, COLING, ACL ARR, and ICASSP.

Outside Research

Outside of work, I follow Formula 1 and travel when I can. Recent trips include Peru, Mexico, Guatemala, Spain, Czechia, Austria, Germany, and Thailand.

After losing about 100 lbs, I got into hiking, running, and powerlifting:

I also volunteer with City Dogs & City Kitties (CDCK), an animal rescue in Washington, D.C.