A

Apu Chandraw Shill

NLP Engineer & AI Researcher

Verbex AI, Dhaka, Bangladesh
B.Sc. in CSE, RUET

News & Updates

Jul 2025 Paper "TituLLMs: A Family of Bangla LLMs with Comprehensive Benchmarking" accepted at ACL 2025 (Findings)
Sep 2023 Joined Verbex AI as NLP Engineer
Nov 2021 Joined BJIT Limited as AI Engineer
Jul 2021 Paper on Plant Disease Detection published at IEEE ACMI 2021
Apr 2021 Joined Prime University as Lecturer, Dept. of CSE
Mar 2021 Graduated from RUET with B.Sc. in CSE (Ranked 9th out of 131)

About Me

I am an NLP Engineer at Verbex AI, where I contribute to innovative projects in natural language processing, leveraging expertise in deep learning and computer vision. My work focuses on developing LLM-based solutions, including expressive semantic translation systems, multi-agent frameworks, and sophisticated RAG pipelines.

I graduated from Rajshahi University of Engineering and Technology (RUET) with a B.Sc. in Computer Science and Engineering, ranking 9th out of 131 students (CGPA: 3.57/4.00). With a commitment to continuous learning, I have completed multiple nanodegrees from Udacity in Deep Learning, Computer Vision, and Data Science.

My research interests span artificial intelligence, with particular focus on large language models, their applications, and security considerations. I am passionate about developing AI solutions that have real-world impact.

Research Interests

  • Large Language Models (LLMs)
  • Natural Language Processing
  • LLM Security & Generative AI
  • Computer Vision
  • Object Detection & Face Recognition
  • Deep Learning
Research Statement

Publications

Conference - ACL 2025

TituLLMs: A Family of Bangla LLMs with Comprehensive Benchmarking

Shahriar Kabir Nahin, Rabindra Nath Nandi, Sagor Sarker, Quazi Sarwar Muhtaseem, Md Kowsher, Apu Chandraw Shill, Md Ibrahim, Mehadi Hasan Menon, Tareq Al Muntasir & Firoj Alam

In Findings of the Association for Computational Linguistics: ACL 2025, pp. 24922-24940. Association for Computational Linguistics. July 2025, Vienna, Austria.

Conference - IEEE 2021

Plant Disease Detection Based on YOLOv3 and YOLOv4

Apu Chandraw Shill, Md. Asifur Rahman

In 2021 International Conference on Automation, Control and Mechatronics for Industry 4.0 (ACMI), pp. 1-6. IEEE. July 2021, Rajshahi, Bangladesh.

Experience

September 2023 - Present

NLP Engineer

Verbex AI

Dhaka, Bangladesh

  • Contributed to an LLM-based Expressive Semantic Translation (EST) system for Bengali, surpassing Google, GPT-4o, IndicTrans2, and Gemini-1.5-pro in semantic preservation
  • Developed collaborative multi-agent systems to generate high-quality synthetic data for LLM applications
  • Fine-tuned LLMs and developed voice bots for industry-specific tasks
  • Designed sophisticated RAG pipelines with optimized retrieval and embedding methods, increasing factual consistency and reducing hallucinations
  • Developed end-to-end pipeline for Bengali intent classification in MFI domain, achieving 93% F1-score
November 2021 - August 2023

AI Engineer

BJIT Limited

Dhaka, Bangladesh

  • R&D on various object detection models (YOLO, Mask R-CNN)
  • Added classification layer on Facenet achieving 98% face classification accuracy
  • Built face recognition systems with VGG-Face, Facenet, ArcFace, InsightFace
  • Implemented modified object/face tracking with SORT algorithm
  • R&D, built and analyzed custom emotion detection models
  • Achievement: 1st place in training program
April 2021 - October 2021

Lecturer, Dept. of CSE

Prime University

Dhaka, Bangladesh

  • Conducted theory and lab sessions for undergraduate students
  • Coordinator: Prime University Programming Club (PUPC)
  • Mentored students in programming and career development
2019 - 2020

Undergraduate Research Assistant

RUET

Rajshahi, Bangladesh

  • Conducted research in computer vision and deep learning
  • Applied YOLO object detection models on real-world data
  • Collaborated with faculty supervisor, leading to peer-reviewed conference publication

Education

January 2016 - March 2021

B.Sc. in Computer Science and Engineering

Rajshahi University of Engineering and Technology (RUET)

Rajshahi, Bangladesh

  • CGPA: 3.57/4.00
  • Rank: 9th out of 131 students
  • Courses: Data Structures, Algorithms, DBMS, Operating Systems, Computer Networks, Artificial Intelligence

Certifications

Deep Learning Specialization

Coursera

Natural Language Processing Specialization

Coursera

Machine Learning

Coursera

Computer Vision Nanodegree

Udacity

Data Scientist Nanodegree

Udacity

TensorFlow in Practice Specialization

Coursera

Skills

Core Technologies

Artificial Intelligence Machine Learning Deep Learning Computer Vision NLP Large Language Models

Programming Languages

Python C C++ SQL MATLAB

LLM Frameworks

OpenAI Agents LangChain LangGraph LlamaIndex AutoGen

ML/DL Libraries

PyTorch TensorFlow NumPy Pandas Scikit-Learn SciPy

Databases

MySQL MongoDB Google BigQuery

Tools & Platforms

Flask FastAPI GCP AWS Git Docker Kubernetes

Selected Projects

Named Entity Recognition System

Built a NER system with custom neural model using Embedding vectors, Bi-LSTM, and CNN on CoNLL2003++ dataset.

Python PyTorch Bi-LSTM

Bengali Person Name Extractor

Developed a system to extract Bengali person names from given sentences using NLP techniques.

Python NLP Bengali

Facial Recognition System

Built a facial recognition system using OpenFace with high accuracy face detection and identification.

Python OpenFace OpenCV

Image Caption Generator

CNN-LSTM based neural model for image captioning using COCO dataset with ResNet-50 encoder and LSTM decoder.

PyTorch CNN LSTM

Person Re-identification

Implemented person re-identification using Siamese Network for tracking individuals across different camera views.

Python Siamese Network Deep Learning

Face Generator (GANs)

Developed a face generator using Generative Adversarial Networks (GANs) for synthesizing realistic human faces.

Python GANs PyTorch

View More Projects on GitHub

Awards & Honors

1st Place - Training Program

BJIT Limited

Achieved first place among all trainees in the AI/ML training program

Government Scholarships

Dhaka Board, Bangladesh

Awarded scholarships for outstanding achievement at Primary and Secondary levels

Science & Math Olympiad

Divisional & National Level

Participated in Science, Mathematics, and Astro Olympiad competitions

Certified Peer Educator

PIACT, Bangladesh (2012)

Certified upon successful completion of peer education training program

Contact

Get in Touch

I'm always open to discussing new research opportunities, collaborations, or interesting projects. Feel free to reach out!

Phone

+880 1996 408842

Location

East Jurain, Dhaka 1204, Bangladesh

Additional Information

IELTS Score

7.5 (L-7.5, R-8, W-6.5, S-7.5)

Languages

Bangla (Native), English (Professional)

Interests

Traveling, Reading, Creative Writing

Volunteer Work

Somanupatik RUET (2017-Present)