Devanandu Sreenath
Software Engineer with a strong AI/ML specialization — building backend systems, cloud pipelines and LLM-powered applications end to end, from API and infrastructure design to model deployment. Most recently a Software Engineer Intern at Uniphore, working across AWS event-driven architecture, FastAPI services and ML pipeline design.
Profile
Software engineer with a strong specialization in AI/ML — comfortable across the full stack of a system, from backend services and cloud infrastructure to model training and deployment. Built software, data engineering and AI/ML solutions using Python, AWS, FastAPI, Apache Airflow and Docker as a Software Engineer Intern at Uniphore.
Designed and deployed an AWS event-driven architecture spanning S3, EventBridge, Lambda, API Gateway and Power BI — cutting manual operations by 40% and enabling automated data processing and analytics. Comfortable owning a project end to end: stakeholder management, cross-functional collaboration and Agile execution included.
Where I've worked
- Designed and deployed AWS event-driven data pipelines, reducing manual operations by 40%.
- Built ETL workflows for data ingestion, preprocessing, feature engineering and ML pipeline support across 10+ Airflow DAGs.
- Managed containerized AI services using Docker, Kubernetes and GitHub Actions-based CI/CD pipelines.
- Developed FastAPI integrations and collaborated with cross-functional Agile teams to improve system reliability and support production operations.
Selected work
Pipelines, models and real-time systems built during my internship at Uniphore and academic work at College.
Architected a serverless pipeline using Lambda, EventBridge, API Gateway, S3 and Airflow to automate data processing and real-time Power BI analytics, with fault-tolerant workflows and automated retry logic that eliminated manual reporting.
Built end-to-end data automation workflows using Apache NiFi, n8n and Python, integrating with Apache Airflow and AWS S3 to orchestrate automated pipelines supporting ML workflows and business analytics.
Built a real-time AI video analytics pipeline with YOLOv8, PyTorch and OpenCV achieving 90%+ detection accuracy at under 200ms latency, managing the full ML lifecycle through to FastAPI-based deployment.
Designed a real-time pipeline using MediaPipe, OpenCV and NLP to convert sign language gestures into Malayalam text end to end in under 200ms, deployed through a FastAPI REST API.
Built a Retrieval-Augmented Generation pipeline using LangChain, OpenAI APIs, Hugging Face and ChromaDB/Pinecone to deliver context-aware responses, deployed as a scalable Dockerized FastAPI service.
Toolkit
Languages
AI/ML & GenAI
Data Pipelines
Backend & APIs
Cloud
Databases & Tools
Background
Let's build something.
Open to Software Engineer / SDE and AI/ML Engineer roles, and collaborations on backend systems, data pipelines, LLM applications, and real-time ML systems.