Shubham Marshettiwar
Software Developer at Cogniquest AI | Ex- Cerence AI | Python | Flask | FastAPI | Mysql | Docker | Kubernetes | GCP | Java |
Pune Division
3+ years of work experience
Feb 20251 yr 6 mosPresent
Bengaluru
• Built scalable Intelligent Document Processing (IDP) solutions using Python, FastAPI, Flask, and microservices architecture.
• Developed and deployed RAG-based AI pipelines using LLaMA 3 and LlamaIndex for semantic search, document extraction, and summarization from unstructured documents.
• Integrated Google Gemini API for multimodal AI document understanding across invoices, forms, and medical records.
• Designed asynchronous processing pipelines using RabbitMQ and Celery, improving concurrent document processing throughput by 40%.
• Engineered OCR and preprocessing workflows using AWS Textract, GCP Vision OCR, and Google Document AI for high-accuracy data extraction and document quality validation.
• Architected REST APIs and backend services for high-volume document processing systems, reducing overall processing latency by 25%.
• Worked closely with CTO and cross-functional teams to design scalable AI workflows and production-grade backend systems.
• Deployed cloud-native workloads on AWS and GCP using Docker, Kubernetes, Pub/Sub, and Cloud Storage services.
• Improved operational efficiency and reduced cloud processing costs by optimizing OCR and AI inference pipelines.
Jul 20231 yr 3 mosSep 2024
Pune District
• Directing the comprehensive development of an AI-powered in-car voice assistant,
managing the entire project lifecycle from initial conception through to deployment,
while ensuring compliance with client specifications and industry standards.
• Engaging in crafting automation scripts aimed at optimizing system processes,
which led to a notable reduction in manual intervention and a marked improvement
in operational efficiency.
• Partnering extensively with clients to collect and evaluate requirements for
quarterly product updates, guaranteeing that all deliverables not only met but often
surpassed client expectations.
• Executing corrective measures for malfunctioning queries by diligently updating
and refining data for the Natural Language Understanding (NLU) model, which
resulted in enhanced system accuracy.
• Collaborating with cross-functional teams to architect and develop APIs for
seamless integration, utilized Python Flask for efficient data management on
Artifactory, and conducted code reviews and refactoring to enhance the
performance and maintainability of Java codebases.
• Successfully led the end-to-end development of a key project as the sole
developer, resulting in a highly functional AI-driven application that improved user
interaction and satisfaction.
Jan 20237 mosJul 2023
Pune, Maharashtra, India
• Supported the development of an AI-powered voice assistant by actively participating in multiple facets of the project, such as coding, testing, and comprehensive documentation.
• Worked alongside senior developers to collect requirements and execute feature implementations, acquiring significant insights into the intricacies of the software development process.
• Assisted in the integration of third-party APIs, which streamlined data flow and improved the functionality of the AI voice assistant, enhancing overall system performance.
• Participated in daily stand-up meetings and sprint planning sessions, gaining valuable insights into agile project management and team dynamics, which fostered a collaborative work environment.
Shubham’s AI Stack
Education
Bachelor of Engineering - BE, Information Technology
Zeal College of Engineering & Research, Pune
20194 yrs2023
Join Shubham and 10k+ other professionals on
Tal is the app where bosses meet engineers & builders directly, without middlemen
Download Tal Boss