Graduate Software Engineer at Cognizant | Researcher in Linguistics, AI, Computer Vision & Mathematical Computing
I’m passionate about building scalable enterprise software and conducting advanced interdisciplinary research across AI, mathematics, language systems, and computational science.
“Where language, logic, and learning meet — that’s where I work.”
- 🔹 Backend Development with python , django ,streamlit , C#,.NET, ASP.NET MVC, Web API
- 🔹 Database Engineering using SQL Server, T-SQL, SQL Profiler
- 🔹 Enterprise Automation with IBM Datacap & ServiceNow ITSM
- 🔹 Research-oriented programming in Python & C++
- 🔤 Researcher in translation & transliteration systems for 140+ languages
- 🔤 Script mapping, phonetic normalization & cross-script NLP pipelines
- 🔤 Low-resource language modeling & multilingual embeddings
- 🔤 OCR-assisted multi-script text processing
- 🔢 Prime sieve optimization & potential prime detection algorithms
- 🔢 Pattern discovery in prime distributions & special prime classes
- 🔢 Extreme computation: π algorithms, transcendental equations
- 🔢 Hybrid symbolic–neural mathematical reasoning models
- 🧠 LLMs, multimodal learning & intelligent agents
- 👁️ Computer vision for anomaly detection, OCR & classification
- 📄 AI-powered document understanding (forms, handwriting, layouts)
- 🧬 AI in healthcare, bio-inspired computing & digital epidemiology
Python, FastAPI, Flask, Django, PySpark, Streamlit,
C#, ASP.NET Core, EF Core, LINQ, C, C++, Angular
TensorFlow, PyTorch, Keras, OpenCV, scikit-learn,
LangChain, HuggingFace, OpenAI API, LLaMA, DeepSpeed,
MLOps, AIOps, LLMOps, GenAI Agents
NumPy, Pandas, Matplotlib, Seaborn, Regex, BeautifulSoup,
Requests, SQLAlchemy, SymPy
SQL Server, Azure SQL, PostgreSQL, MySQL, Oracle,
MongoDB, NoSQL, PL/SQL, Cosmos DB
Azure: App Service, Functions, Storage, Databricks, AI Studio
AWS: Lambda, Bedrock, SageMaker
GCP: Vertex AI, BigQuery
Docker, Jenkins, Git, Ansible, CI/CD Pipelines
Ubuntu, Debian, RedHat, CentOS, NixOS, Slax OS
Power BI, Tableau, Plotly, Pandas Profiling
ServiceNow, IBM Datacap, BMC Remedy, Jira Service Management, IAM
Python, TensorFlow, Keras, OpenCV, NumPy, Pandas
- Trained LSTM & GRU models on 900+ surveillance videos (arson, burglary, vandalism)
- Used CNN embeddings + temporal modeling for sequence classification
- Achieved 92% accuracy with optimized preprocessing pipelines
Python, Tesseract, spaCy, PyPDF2, Streamlit
- Built ATS system with NER-based skill extraction
- TF-IDF + cosine similarity for JD matching
- OCR support for scanned resumes, improved keyword accuracy by 45%
Python, Streamlit, OCR, Indic Transliteration, OpenAI GPT-4o
- Transliteration across 80+ world scripts with auto script detection
- Context-aware transliteration using LLM + heuristic validation
- Modular design extendable to FastAPI/Django APIs
- Roy, S. (2021) — Video Anomaly Detection with Deep Learning, ResearchGate
- Roy, S. (2022) — Detection of Abnormal Human Behavior Using Deep Learning, Bohr Publications (BIJSCIT)
- Roy, S. (2024) — Crowd Anomaly Detection with Deep Learning, Google Scholar
- 🌍 Ultra-scalable multilingual translators & transliterators
- 🧮 Novel prime-search heuristics & symbolic computation systems
- 🤖 Hybrid symbolic + neural math reasoning architectures
- 📄 High-accuracy OCR & document intelligence pipelines
- 💡 Open to collaboration in AI, Linguistics, Prime Research & Computer Vision
- 🔬 Interested in interdisciplinary and experimental research projects
- 🌱 Always learning, experimenting, and pushing algorithmic limits
📫 Feel free to connect and collaborate!