ARTIFICIAL INTELLIGENCE
AgenticHire
Building AI tools to support first-round technical candidate screening and evidence-based hiring decisions.
ABOUT
Hi, I’m Sabik.
I build robots, explore AI safety, and turn engineering ideas into working systems.
I’m a Grade 12 science student at Bangladesh Air Force Shaheen College Kurmitola. My work brings together robotics, multilingual AI safety, and sustainable engineering. I’m also a co-founder of AgenticHire and Country Director of Young Founders Summit for the 2026 cohort.
Projects
All projectsROBOTICS
Zephyron
A solar-assisted mobile manipulator for environmental sensing and distributed visual inference, with reproducible engineering analysis.
SUSTAINABLE ENGINEERING
BLUESHIELD FILTER
A preliminary floodwater-treatment design combining graphene oxide separation and UV-C treatment, supported by engineering calculations.
ROBOTICS
A3P5 NEMESIS
A rover design for environmental observation and robotic sampling, with reproducible mobility analysis and sensor-calibration research.
Research work
All research
A Hybrid TFT-NBEATS Framework for Robust Multi-Horizon Photovoltaic Power Forecasting
arXiv preprint19 October 2026
An attention-fused TFT and N-BEATS model for solar-power forecasting across five horizons, from 30 minutes to 24 hours, evaluated on four DKASC datasets.
View detailsHide details for A Hybrid TFT-NBEATS Framework for Robust Multi-Horizon Photovoltaic Power Forecasting
5forecast horizons
4DKASC datasets
2.88–4.02%reported MAPE
Abstract summary
This study combines the Temporal Fusion Transformer’s handling of short-term changes with N-BEATS’ trend and seasonality modelling. Learned attention weights join the two branches to forecast photovoltaic output at five horizons. Evaluation uses four periods of DKASC measurements, with comparisons against statistical, machine-learning, and neural baselines. Attention visualisations, decomposition, and SHAP help explain predictions; a Hugging Face deployment accompanies the paper.
30 min6 steps
2 hours24 steps
6 hours72 steps
12 hours144 steps
24 hours288 steps
| DKASC period | RMSE | MAE | MAPE (%) |
|---|---|---|---|
| 2016–2017 | 42.35 | 35.41 | 2.88 |
| 2023 | 46.27 | 38.00 | 3.29 |
| 2024 | 52.29 | 40.08 | 4.02 |
| 2025–2026 | 48.11 | 39.61 | 3.91 |
Each period was trained and evaluated independently. Values are reported by the authors in the linked preprint.
Authors & affiliations
- Shafi Bin Sultan1†St. Joseph Higher Secondary School, Dhaka, Bangladesh
- Sabik Bin Sultan2†Bangladesh Air Force Shaheen College, Kurmitolla, Dhaka, Bangladesh
- Mohammad Tajrian Arafat Alvi3Computer Science & Engineering, Jahangirnagar University, Savar, Bangladesh
- Rayhanul Islam Sony4Trine University, Angola, Indiana, USA
- Abdulah Al Numan5Sembcorp North-West Power Company Ltd, Dhaka, Bangladesh
† Shafi Bin Sultan and Sabik Bin Sultan contributed equally.
Corresponding author: Abdulah Al Numan.
Dataset, methodology & comparisons
- Source
- Desert Knowledge Australia Solar Centre (DKASC), Alice Springs, Northern Territory, Australia
- Sampling
- Every 5 minutes
- Target
- Active power output
- Inputs
- Global horizontal irradiance (GHI), diffuse horizontal irradiance (DHI), ambient temperature, relative humidity, and wind speed
- Split
- Chronological: 70% training / 15% validation / 15% test
- Input window
- 144 observations (12 hours)
| Period | Train | Validation | Test |
|---|---|---|---|
| 2016–2017 | 147,743 | 31,660 | 31,659 |
| 2023 | 71,331 | 15,286 | 15,285 |
| 2024 | 71,331 | 15,286 | 15,285 |
| 2025–2026 | 73,770 | 15,808 | 15,808 |
Preprocessing
- Missing-value imputation
- Kalman filtering
- Isolation Forest outlier detection
- Z-score filtering
- Wavelet denoising
- RobustScaler
Model architecture
- TFT branch
- Short-term dependencies and feature selection
- N-BEATS branch
- Trend and seasonality decomposition
- Attention fusion
- Adaptive weighting of both branches
- Interpretability
- Attention visualisation, trend–seasonality decomposition, and SHAP
Comparison models
ARIMA · SARIMA · Random Forest · XGBoost · LSTM · GRU · Transformer · N-BEATS · TFT
The ablation study compares TFT only, N-BEATS only, fusion without attention, and the full hybrid model (Table 11).
Training configuration & future work
- Batch size
- 64
- Hidden dimension
- 64
- TFT LSTM layers / attention heads
- 1 / 4
- N-BEATS hidden size
- 128
- Optimiser
- Adam
- Learning rate
- 10⁻³ (fusion), then 10⁻⁴ (end-to-end)
- Weight decay
- 10⁻⁵
- Loss
- Quantile + Huber
- Scheduler
- ReduceLROnPlateau
- Epochs
- 25 per phase
- Early-stopping patience
- 25 epochs
- Compute
- Google Colab Premium · NVIDIA Tesla T4 (16 GB)
- Software
- Python 3.10 · TensorFlow/Keras · Scikit-learn · Pandas · NumPy
Future work
Future directions include probabilistic forecasts, numerical weather prediction inputs, and more efficient large-scale deployment.
Funding: No specific grant reported.
Publication details & citation
- Status
- Preprint · Submitted to Elsevier
- Submitted
- 19 October 2026
- Version
- 2610.27155v1
- Category
- Machine Learning (cs.LG)
- Length
- 17 pages
- DOI
- 10.48550/arXiv.2610.27155
Cite this paper
Shafi Bin Sultan, Sabik Bin Sultan, Mohammad Tajrian Arafat Alvi, Rayhanul Islam Sony, Abdulah Al Numan. (2026). A Hybrid TFT-NBEATS Framework for Robust Multi-Horizon Photovoltaic Power Forecasting. arXiv:2610.27155.
Download BibTeX ↗Publication and study details are based on the arXiv v1 manuscript.
Photovoltaic forecastingMulti-horizon predictionHybrid deep learningTemporal Fusion TransformerN-BEATSExplainable AI
Register Shifts Break LLM Safety: A Bengali Benchmark with Culturally Grounded Harms
Findings of EMNLP 20262026
A Bengali safety benchmark of 879 prompts across 17 culturally grounded harm categories, evaluating how language, register, and framing affect 18 LLMs.
View detailsHide details for Register Shifts Break LLM Safety: A Bengali Benchmark with Culturally Grounded Harms
Research summary
A Bengali safety benchmark of 879 prompts across 17 culturally grounded harm categories, evaluating how language, register, and framing affect 18 LLMs.
- Publication
- Findings of EMNLP 2026
- Status
- Accepted
- arXiv submission
- 23 August 2026
- Authors
- Naymul Islam, Nusrat Jahan Lia, Shubhashis Roy Dipta, Sabik Bin Sultan, Abdullah Khan Zehady
Zephyron: Integrated Design and Analytical Evaluation of a Solar-Assisted Mobile Manipulator for Multimodal Environmental Reconnaissance and Distributed Visual Inference
arXiv preprint2026
An integrated design and analytical evaluation of a solar-assisted rover combining environmental sensing, manipulation, and distributed visual inference.
View detailsHide details for Zephyron: Integrated Design and Analytical Evaluation of a Solar-Assisted Mobile Manipulator for Multimodal Environmental Reconnaissance and Distributed Visual Inference
Research summary
An integrated design and analytical evaluation of a solar-assisted rover combining environmental sensing, manipulation, and distributed visual inference.
Study scope
Analytical scenarios and design proposals; paper does not establish experimental payload, endurance, or detection ratings.
- Publication
- arXiv preprint
- Status
- Preprint
- arXiv submission
- 22 September 2026
- Authors
- Sabik Bin Sultan, Shafi Bin Sultan, Safwan Sadad
A3P5 NEMESIS Integrated Rover Design for Environmental Reconnaissance and Robotic Sampling with Reproducible Mobility Analysis and an External Data Machine Learning Calibration Benchmark
arXiv preprint2026
A rover design study combining environmental reconnaissance, robotic sampling, reproducible mobility analysis, and an external-data calibration benchmark.
View detailsHide details for A3P5 NEMESIS Integrated Rover Design for Environmental Reconnaissance and Robotic Sampling with Reproducible Mobility Analysis and an External Data Machine Learning Calibration Benchmark
Research summary
A rover design study combining environmental reconnaissance, robotic sampling, reproducible mobility analysis, and an external-data calibration benchmark.
Study scope
Integrated field performance remains unverified; external-data sensor calibration results do not establish rover accuracy.
- Publication
- arXiv preprint
- Status
- Preprint
- arXiv submission
- 16 September 2026
- Authors
- Shafi Bin Sultan, Sabik Bin Sultan, Safwan Sadad
Experience
AgenticHireSINGAPORE · ENTREPRENEURSHIP
Young Founders Summit
Building AgenticHire with young founders from different countries through mentorship and venture development.
Read my reflection on LinkedInWork
All workLet’s connect.
Research, engineering, and building things together.