Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Software Engineer with 4+ years of experience building Python back-end and AI/ML systems, specializing in LLM applications, semantic search, and cloud-native architectures on AWS and Azure. Experienced in production ML with strong MLOps practices and scalable API-driven solutions. I build and evaluate reliable AI/LLM pipelines—covering RAG, prompt engineering, intent classification, and model safety/regression testing—while also delivering data pipelines, observability, and CI/CD automation to improve performance and reduce deployment friction.…Software Engineer with 4+ years of experience building Python back-end and AI/ML systems, specializing in LLM applications, semantic search, and cloud-native architectures on AWS and Azure. Experienced in production ML with strong MLOps practices and scalable API-driven solutions. I build and evaluate reliable AI/LLM pipelines—covering RAG, prompt engineering, intent classification, and model safety/regression testing—while also delivering data pipelines, observability, and CI/CD automation to improve performance and reduce deployment friction.

SHREYA THAKUR

Software Engineer, AI Engineer, Back-End Developer, +2





Software Engineer with 4+ years of experience building Python back-end and AI/ML systems, specializing in LLM applications, semantic search, and cloud-native architectures on AWS and Azure. Experienced in production ML with strong MLOps practices and scalable API-driven solutions. I build and evaluate reliable AI/LLM pipelines—covering RAG, prompt engineering, intent classification, and model safety/regression testing—while also delivering data pipelines, observability, and CI/CD automation to improve performance and reduce deployment friction.…Software Engineer with 4+ years of experience building Python back-end and AI/ML systems, specializing in LLM applications, semantic search, and cloud-native architectures on AWS and Azure. Experienced in production ML with strong MLOps practices and scalable API-driven solutions. I build and evaluate reliable AI/LLM pipelines—covering RAG, prompt engineering, intent classification, and model safety/regression testing—while also delivering data pipelines, observability, and CI/CD automation to improve performance and reduce deployment friction.

Available to hire

Software Engineer with 4+ years of experience building Python back-end and AI/ML systems, specializing in LLM applications, semantic search, and cloud-native architectures on AWS and Azure. Experienced in production ML with strong MLOps practices and scalable API-driven solutions.

I build and evaluate reliable AI/LLM pipelines—covering RAG, prompt engineering, intent classification, and model safety/regression testing—while also delivering data pipelines, observability, and CI/CD automation to improve performance and reduce deployment friction.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Intermediate

Intermediate

Intermediate

Intermediate

Intermediate

Language

Work Experience

AI Software Developer at Sayaam for All

August 1, 2025 - Present

Engineered LLM-powered backend services using Python, FastAPI, and REST APIs, integrating RAG with Pinecone to enable scalable inference and improve response efficiency by 40%. Designed a zero-shot intent classification pipeline using Hugging Face BART-MNLI for dynamic routing, improving classification accuracy by 25%. Containerized backend and ML services using Docker and CI/CD and deployed on AWS EC2, reducing setup time by 70%.

Graduate Teaching Assistant at University at Buffalo

September 1, 2024 - June 30, 2025

Mentored 200+ students in deep learning and data-intensive computing covering CNNs, Transformers, and model evaluation. Supported applied research on an end-to-end document intelligence system (OCR, summarization, translation, and TTS). Architected a custom OCR pipeline using ResNet, CRNN, BiLSTM, and CTC with ~92% accuracy and ~35% CER reduction, using GPU inference and a Docker-based Azure deployment (ACR, Web App).

Software Engineer at Wipro Ltd.

August 1, 2019 - August 1, 2022

Built an AI-powered semantic search platform using Python, Azure Cognitive Search, and NLP, indexing 10M+ documents and improving search precision by 25% while reducing query latency by 40%. Developed ML-based PII detection using Flask, Presidio, and event-driven Azure Functions, reducing compliance effort by 80%. Implemented an NLP preprocessing microservice in Python/NLTK improving ML accuracy by 20% via tokenization and normalization. Designed analytics ingestion pipelines with Azure Blob Storage and SQL for batch processing 10TB+ data, reducing latency by 35%. Improved CI/CD reliability using Docker, Azure DevOps, and Key Vault, reducing deployment failures by 30%. Also built Cisco Partner Portal backend services in Java with SQL and REST APIs, supporting partner onboarding and reducing API latency by 35%, and automated testing with Selenium/Cucumber to achieve 95% test coverage.