SANKARDEV S
AI/ML Engineer with 3+ years building production ML systems. Published researcher with 95.08% F1-score NLP model (JDSIS 2026). Specialising in LLMs, NLP pipelines, and multimodal AI — targeting AI roles and industrial PhD opportunities in Germany.
model = RoBERTa.from_pretrained(
"emotion-classifier"
)
result = model.predict(text)
▶ F1-Score: 95.08% · Classes: 28
🎓 Research Interests · Industrial PhD
Research Metrics
F1-Score
RoBERTa NLP · JDSIS 2026
Text Samples
GoEmotions + Hate Speech pipeline
Latency Reduced
ML training optimisation
CREMA-D Accuracy
Multimodal Wav2Vec2.0 model
Emotion Classes
EmoTract real-time classifier
Technical Stack
Built for AI at Scale
The Journey
Professional milestones and academic achievements.
Research Assistant
- Built a data pipeline processing 250,000+ text samples from GoEmotions and Hate Speech corpora.
- Fine-tuned RoBERTa for multi-class emotion recognition; achieved 95.08% F1-score, published in JDSIS (2026).
- Reduced model training latency by 30% through batching optimisations across classical and transformer-based models.
- Developed SoundSense, a multimodal vocal anomaly detection system using Wav2Vec2.0.
- Achieved 78.10% accuracy on CREMA-D, outperforming SVM/RF baselines by 30%.
Software Engineer
- Built backend systems using Laravel and MySQL for multiple concurrent client projects.
- Delivered React.js frontends with full mobile responsiveness across all deliverables.
Software Associate
- Delivered 200+ bespoke software solutions across Python, PHP, and React; 95% client satisfaction.
- Deployed predictive ML models into production, reducing manual data processing time by ~20%.
Education
M.Sc. Artificial Intelligence (with Industrial Placement)
2023 – 2025University of East London, UK
B.Sc. Computer Applications
2019 – 2022Kerala University, India
Selected Projects
Things I've Built
VoxChain – Real-Time LLM Voice Agent
Real-time voice AI system with bidirectional streaming (speech ↔ LLM ↔ audio). Deployed end-to-end agentic LLM pipeline with low-latency audio I/O.
SmartSafe – Multimodal Liveness Detection
Attendance system using facial recognition combined with audio liveness scoring for anti-spoofing. Generates a fraud-resistance confidence score from vision and audio signals.
SoundSense – Multimodal Emotion Detection
Cross-modal system detecting emotional inconsistencies between speech text and vocal tone. Achieved 78.10% accuracy, outperforming SVM/RF baselines by 30%.
Latest Publication
"NLP-Framework for Youngsters with Advanced Transformer-Based Models"
Journal of Data Science and Intelligent Systems (JDSIS), 2026
Let's Connect
I'm currently open to new opportunities in AI Research and Full-Stack Engineering. Whether you have a question or just want to say hi, I'll try my best to get back to you!
© 2026 Sankar Dev S. All rights reserved.