Apr 2025 – Mar 2026 · 11 mos · Pune · Hybrid

• This is a JV between Accenture and Telstra
• Designed, Built and Solved the GenAI Call Drivers automation, a high profile transcript analysis tool that identifies root causes and product-level issues using GPT-4.1-mini. I am the tech lead of the team where I mentored and built this solution which is able to give insights to the lowest level of customer issues that they mention over IVR call. It compiled a comprehensive daily report across the Telstra product suite: 25K calls a day are analyzed within 3 hours. This insight report is passed to the board of directors and business stakeholders who can now fix business processes and can earn and save the business multimillion dollars. The solution I built is an end to end daily pipeline between Azure Data Factory, AzureML and PowerBI which collects data, transforms it, processes the call transcripts and publishes the results in a powerBI dashboard for the business executives. Reasoning was built into the prompt to get the maximum benefit of the GPT-5-mini reasoning model. I have built all parts of this pipeline.
• Built and enhanced AskTelstra GenAI features, AskTelstra is a production grade RAG based Langchain + LangGraph GenAi Chatbot. The self service architecture I built has reduced build costs by 88%, improved user experience, and enabled scalable deployment to over 15,000 frontline agents. Created the multiple promotions query flow, undertook retrieval, indexing and chunking experiments to improve the response quality of the chat-bot.
• Built the GenAI ENPS verbatim analyser, a GenAI workflow tool using GPT 4.1 mini that extracts
• sentiment and themes from employee feedback, saving a month of manual effort and potentially improving ENPS scores by 2 points.
• Leader and Mentor to developers, accelerating onboarding, development and enabling faster contributions

Nov 2024 – Jun 2025 · 7 mos

• I have successfully automated and updated the NATAMA Fault Data Process, ensuring compatibility with Azure ML by replacing the existing Chroma DB with FAISS DB and addressing data formatting issues. This was completed before the year-end deadline, with permanent steps in place to resolve data formatting issues.
• I mentored developers and improved the onboarding process, conducting sessions to walk them through platforms, tools, and the codebase. This led to efficient onboarding and faster contributions to the project. I have also mentored around design, architecture and software best practices.
• Developing an ML model to identify PII in SMS texts, I used the XGBoost model for experimentation. The model performed exceptionally well, improving data security for customer communications.
• Additionally, I built the Admin Portal for AskTelstra V2, allowing business units to manage configurations and ensuring seamless updates across multiple units.
• I also built a prompt portal allowing business units to update and modify prompts on the fly without redeploying the entire application.

May 2024 – Nov 2024 · 6 mos

• As an AI/ML Engineer, I played a crucial role in enhancing system architecture, security, and performance for AskTelstra. My work primarily focused on:
• Unit Testing & Code Quality: Introduced unit testing with pytest, improving error handling, stability, and bug tracking.
• Security & Authentication: Implemented backend authentication and SSO using JWT tokens, ensuring secure user access and facilitating user profile integration.
• Logging & Data Analysis: Designed a comprehensive logging framework with unique identifiers (topic_id, interaction_id, dependency_id) for better searchability, analysis, and troubleshooting.
• Operational Efficiency & Deployment: Streamlined release processes with a fast-paced branching and merging strategy for weekly releases, successfully deploying AskTelstra to 1,500 frontline agents with plans to scale to 5,000.
• Admin Portal Ownership: Enhanced user management through role-based access control, bulk user uploads, and configuration management.
• Team Collaboration & Knowledge Sharing: Fostered a collaborative culture by conducting onboarding, mentoring, and delivering knowledge-sharing sessions.
• Performance Optimization: Improved system performance by modularizing the codebase, reducing file sizes, optimizing chunking and indexing, and significantly reducing request latency.
• Scalability & Business Impact: Contributed to the backend architecture to support scaling, ideating and implementing solutions that addressed business challenges, including custom data dashboards and compliance with versioning mandates.
• Recognized with a Team Award for 2024 Q4 and 2025 Q1 for my contributions. Also received the Individual Award - Data-engineering India Achiever of the Month, January 2025.

Oct 2022 – Apr 2024 · 1 yr 6 mos

• ML Developer
• Mapped a scalable pathway to onboard new NLP Mailboxes within timelines for stakeholders
• Tested and Implemented several ML and DL algorithms from reknown ML Libraries to build email segregation systems from stages of Data cleaning to ML pipelines
• Based on the problem statement, went through different stages of ML development including but not limited to Data Cleaning,
• Data Analysis, ML Modelling, ML pipelining, Documentation and presentation.
• Java Developer
• Took ownership of the API development platform as the lead developer.
• Implemented connection pooling for the APIs that improved latency and response time by over 500%
• Upgraded APIs and increased code coverage to atleast 95%, to resolve tech debt of the API space making the APIs more robust, efficient and secure.
• Mentored Junior developers on the platform on how to develop fault tolerant APIs.

Jul 2021 – Oct 2022 · 1 yr 3 mos

• Chatbot Developer
• Delivered chat flows with automated testing based on stakeholder requirements
• Developed a Google widget that saves 50% time in development time and testing.
• Developed an Alert system that can alert a user of an update in chat flows.
• Developed Proof-of-concept Natural Language Processing Models to
• 1. Used Sentiment analysis to Assimilate whether a work completed by a worker was positive or negative.
• 2. Used topic modeling to determine where the problem exists; if multiple users are facing a similar issue.
• Java Developer
• Developed scalable springboot APIs after conversing with BAs and deployed them on Azure instances.
• Worked on Azure functions PoC with senior developers and helped in Junit5 migration.

Jan 2021 – Apr 2022 · 1 yr 3 mos · Mumbai, Maharashtra, India

• Content Creator for Soft skills and Aptitude.
• Headed various meetings with Colleges to aid in their Job opportunities.
• Content writer for various posts and updates of the company.
• Brainstormed several ideas and implemented them to aid the growth of the business.

May 2019 – Jul 2019 · 2 mos · Mumbai, Maharashtra, India

Dec 2018 – Jan 2019 · 1 mo · Mumbai, Maharashtra, India

• Intern at COE-CNDS, working on Machine Learning.