Kushagra Agrawal

Prof. Dr. Kushagra Agrawal

Department of Computer Science and Engineering, Manav Rachna International Institute of Research and Studies, Faridabad, India
orcid

Research Focus & Metrics

Academic Career Length

1Year (2026-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 January 2019 in International Journal of Information Dissemination and Technology
Verified
Journal Article
Closed access
PDF file
Kushagra AgrawalKushagra Agrawal
Kushagra Agrawal
NKNikunj Kamboj
Nikunj Kamboj
In India collecting information like fertility of the soil, weather, growth of crops, temperature, rainfall and information regarding plantation of seeds, etc. can be collected with the help of IOT. It helps the farmers to gain information regarding all agriculture activities. With the help of internet technology, agriculture processes can be monitored through sensors, smart cameras, mobile applications and devices like mini chips. Through IOT the automated technology helps the farmers infrequent ways i.e. best utilization of resources as resources are very scared and solving their agriculture issues from finding the proper time of seeding to the cutting of crops.

30 April 2023
Verified
Book Chapter
Closed access
PDF file
Kushagra AgrawalKushagra Agrawal
Kushagra Agrawal
RKRajneesh Kumar
Rajneesh Kumar
SJShaveta Jain
Shaveta Jain


Verified
Journal Article
Open access
PDF file
Kushagra AgrawalKushagra Agrawal
Kushagra Agrawal
NDNirmiti Deshmukh
Nirmiti Deshmukh
PKPalak Kaushik
Palak Kaushik
SJShaveta Jain
Shaveta Jain
GSGanga Sharma
Ganga Sharma
SYSumendra Yogarayan
Sumendra Yogarayan
Large language models (LLMs) incur prohibitive computational costs when deployed as monolithic systems for multi-domain query processing. This paper proposes Accelerated LLM, a modular architecture that replaces a single general-purpose LLM with an ensemble of task-specialised small language models (SLMs) governed by a neural query router and a Mamdani fuzzy inference system. The router embeds each user query using a frozen sentence encoder and classifies it across four task domains—summarisation, translation, question answering, and text generation—routing confident queries directly to the corresponding SLM. Ambiguous queries are escalated to a three-input fuzzy logic system operating on Query Length, inter-Domain Overlap Score, and Classifier Confidence, enabling principled handling of imprecise inputs. A reinforcement-learning feedback loop, validated through a controlled pilot deployment, continuously refines the routing policy. The complete pipeline, including the sentence encoder, totals approximately 2.14 billion parameters—a 98.8% reduction relative to GPT-3.5 (175 B). The integration of fuzzy logic into the routing stage raises classification accuracy from 91.5% to 94.3% and reduces the hallucination rate to 9.8% (minor) and 6.4% (major). Evaluated on healthcare-augmented benchmarks against ChatGPT-3.5, Claude, Mistral 70B, and two contemporary compact models (GPT-4o-mini and Llama 3.1-8B-Instruct), Accelerated LLM achieves competitive or superior task-specific performance at a fraction of the parameter count. A small-scale pilot evaluation in the legal domain indicates that the routing and fuzzy logic components retain partial effectiveness beyond the primary healthcare setting, though full multi-domain validation remains future work.