Machine Learning Ai Engineer

Zeeshitha Buridi

+1-704-***-**** ************@*****.*** Linkedin

SUMMARY:

Certified Azure AI Engineer with expertise in prompt engineering using Azure AI Studio/Foundry, optimizing GPT-3, GPT-4, and implementing advanced prompt flows (RAG, dynamic prompting, multi-model workflows).

6+ years of experience as an AI Engineer/Conversational Designer, specializing in chatbot development and NLP-driven user experiences.

Proficient in designing and deploying chatbots using Avaamo, Microsoft Bot Framework, Dialogflow CX, and GCP, optimizing functionality and enhancing user engagement.

Strong knowledge of state-based conversation design, intent handling, entity extraction, session management, and webhook integration in Dialogflow CX.

Experienced in Generative AI, Azure Machine Learning, AutoML, and MLOps for model development, training, and deployment.

Skilled in Google Cloud Functions, Cloud Run, Python Flask webhooks, and Google Cloud Logging for cloud-based chatbot interactions.

Developed backend services and APIs for seamless chatbot integration with web and mobile applications, enhancing user experience.

Led end-to-end chatbot development, integrating REST/SOAP APIs, deploying on SharePoint, Skype, MS Teams, and implementing CI/CD pipelines for efficient deployments.

Proficient in JavaScript, Python, React, and Node.js, with expertise in NLP frameworks like TensorFlow and BERT for intent recognition and sentiment analysis.

Collaborated with UX/UI teams to design intuitive chatbot interfaces, optimizing usability and engagement.

Experience in LivePerson and Kore.ai, ensuring chatbot adaptability through advanced NLP techniques (Azure Cognitive Services, Sentiment Analysis, Entity Recognition, OCR, Language Translation).

Developed PoC chatbots for MSDP field engineers, integrating incident tracking with real-time status updates via RESTful APIs in Node.js, improving operational efficiency.

CERTIFICATIONS:

Microsoft Certified: Azure AI Engineer Associate Issuing Organization: Microsoft

Date: October 2024

Credential ID: 19A2D3070425A20C CREDENTIAL URL

TECHNICAL SKILLS:

Chatbot Development & NLP

Avaamo, LivePerson, Kore.ai, Dialogflow, Microsoft Bot Framework, TensorFlow, BERT, IBM Watson, Google Dialogflow CX, Kore.ai XO Platform

Prompt Engineering for LLMs (GPT-3, GPT-4, Claude, Gemini)

SOAP Notes Generation, GPT-4, spaCy NLP, OpenAI API, Named Entity Recognition (NER), Text Classification, Data Preprocessing, API Integration,

Clinical NLP Automation

Programming & Web Technologies

JavaScript (Avaamo, SharePoint, Skype, MS Teams), HTML, CSS, Python, Node.js, ReactJS, TensorFlow, PyTorch, Hugging Face Transformers

Python NLP Pipeline, Medical Transcript Processing, RegEx, Named Entity Recognition (NER), Batch Text Chunking, Structured Data Extraction

API Integration & Backend Development

REST API, SOAP API Integration, Google Cloud Functions, Python Flask Webhooks

SOAP Notes Automation, API-Based Text Classification, OpenAI GPT-4 API, SOAP Framework (Subjective, Objective, Assessment, Plan), NLP-Based Medical Text Processing

Cloud

Azure, Cloud Platform (GCP)

Azure AI Services (NLP Techniques)

Text Analytics, Language Understanding, Sentiment Analysis, Natural Language Processing, Entity Recognition, Language Translation, Optical Character Recognition (OCR), Content Moderation, Document Intelligence, Anomaly Detection

Azure AI Studio/Foundry (Prompt Engineering with GPT-3 and GPT-4), Azure OpenAI models, LLM (Large Language Models) via Azure OpenAI, RAG (Retrieval-Augmented Generation) using Azure Cognitive Search and OpenAI, Azure OpenAI Embedding models(text-embedding-ada-002, text-embedding- 3-small, text-embedding-3-large), Vector databases

Azure Application and

Integration Services

Azure App Service, Azure Functions, Azure Logic Apps

Azure Machine Learning & JupyterLab

Azure ML JupyterLab for developing, training, and deploying machine learning models

Named Entity Recognition (NER), GPT-4 Prompt Engineering, Medical Text Processing, OpenAI API, Clinical Documentation Automation

Azure Machine Learning: Model Development, Model Training, AutoML, Model Deployment, Predictive Analytics, Scalable Machine Learning, ML Operations (MLOps)

Generative AI: Experience with AI model generation and fine-tuning

Voice Technology &

Speech Application

Azure Speech Services: Speech-to-Text (STT), Text-to-Speech (TTS), Speech

Translation

Interactive Voice

Design & IVR Engine

Azure Communication Services, Speech Service, Twilio

LLM Fine-tuning & Code Generation

Experience with fine-tuning LLMs for code generation (Python, JavaScript, SQL)

AI Model Fine-tuning and contextual programming

Version Controls

GitHub, Jenkins, CI/CD Implementation

Natural Language

Processing (NLP) and Machine Learning (ML)

NLP, Deep learning, Model fine-tuning, Reinforcement learning

Azure Storage and

Database Systems

Azure Blob Storage, SQL, MongoDB

Operating Systems

Windows

Robotic Process

Automation

Power Automate

IDE and Tools

Microsoft Visual Studio

MS Software Packages

MS Office (Word, Excel, PowerPoint)

Project Management

JIRA, BMC Service Broker, BMC ITSM, SAP, MWP & PUC Products

PROFESSIONAL EXPERIENCE:

Client: PCI Feb 2023 to till date

Location: New Jersey

Project: AI-Powered Virtual Assistant for Smart Infrastructure Support (IVR + Chatbot with RAG) Role: Senior Azure AI Engineer

Summary: Designed and implemented an AI-powered IVR and chatbot solution to support PCI’s smart infrastructure services by automating incident reporting, device diagnostics, and customer support ticketing. The system, powered by Retrieval-Augmented Generation (RAG) and Microsoft Azure AI services, enabled natural, real-time support interactions across voice and chat channels. Integrated with PCI’s IoT platform, CRM system, and Microsoft Teams, the assistant reduced manual effort, streamlined field technician workflows, and enhanced the efficiency of operations across smart city and building management

environments.

Key Responsibilities:

Developed and deployed a virtual assistant with Retrieval-Augmented Generation (RAG) to retrieve real-time insights from PCI’s IoT systems, knowledge bases, and documentation.

Created intelligent IVR flows using Twilio’s programmable voice and Azure Speech Services (STT/TTS) to support phone-based incident reporting for building operators.

Built context-aware chatbot flows using Microsoft Bot Framework and deployed them

on Microsoft Teams, enabling field staff to access diagnostics, reports, and submit service requests.

Fine-tuned GPT-3 and GPT-4 responses via prompt engineering in Azure AI Studio, improving interaction quality and relevance for issues like HVAC malfunctions, sensor alerts, and energy metrics.

Developed a Python-based content intelligence assistant leveraging prompt chaining techniques to extract, summarize, and classify external IoT documentation. Used multi-stage prompts to identify relevant links, summarize content, and generate structured outputs in Markdown or JSON.

Engineered contextual system/user prompts to guide GPT behavior in link extraction, instruction summarization, and multilingual content generation—supporting PCI's global operations with localized responses.

Applied prompt optimization strategies, including example-based prompting and token-aware truncation, to ensure high-quality, structured, and consistent LLM outputs for internal and field support usage.

Designed and integrated Python-based APIs and webhooks to enable seamless data flow between the bot, PCI’s CRM, and backend IoT ticketing systems.

Automated support workflows using Power Automate, including ticket creation, incident escalation, system alerts, and real-time notifications.

Conducted pre-deployment testing using Bot Emulator and post-deployment monitoring

via Azure Application Insights, enabling iterative improvements driven by telemetry and user feedback.

Improved operational efficiency by automating 60% of routine support tasks, handling over 500+ monthly interactions, and reducing manual workload for the PCI support team by 40%.

Technologies:

LLMs & RAG: GPT-3, GPT-4, Retrieval-Augmented Generation (RAG), Prompt Chaining, System/User Prompts

Microsoft Azure: Azure AI Studio, Azure Speech Services (STT/TTS), Azure LUIS, Azure Functions, Azure Application Insights

Conversational Platforms: Microsoft Bot Framework, Microsoft Teams

Voice Integration: Twilio (Programmable Voice, Webhooks)

Automation & Integration: Power Automate, Python (API integration, NLP, scraping logic with BeautifulSoup)

Content Intelligence: OpenAI SDK, JSON/Markdown formatting, Multilingual Translation (e.g., Spanish), Role-based Prompting

Project: AI-Powered SOAP Notes Generation System Role: AI Engineer NLP Developer

Summary: Developed an AI-driven clinical documentation assistant that automates SOAP notes generation using GPT-4 and spaCy. The system processes doctor-patient conversations, classifies text into SOAP sections

using NLP-based entity recognition, and generates structured clinical notes. Integrated OpenAI API for LLM- based text summarization and classification, optimizing medical documentation workflows. Designed

a scalable, automated pipeline for handling medical transcripts efficiently.

Responsibilities:

Designed and implemented an AI-powered text processing pipeline to convert doctor-patient transcripts into structured SOAP notes.

Developed Named Entity Recognition (NER)-based classification using spaCy, mapping conversations into Subjective, Objective, Assessment, and Plan (SOAP) sections.

Implemented text preprocessing techniques using Python (RegEx, tokenization, and chunking) for cleaning and structuring transcripts.

Integrated GPT-4 API to generate professional SOAP notes, ensuring clarity, coherence, and medical documentation standards.

Built a secure API integration pipeline using OpenAI’s API, handling authentication via. env and dotenv.

Optimized batch text processing to handle large transcripts efficiently, reducing token consumption and improving response time.

Designed an error-handling mechanism for invalid API responses, ensuring system reliability.

Deployed and tested the workflow on local and cloud-based environments to validate accuracy and efficiency.

Automated the end-to-end process for extracting, structuring, and generating SOAP notes, reducing manual effort for clinical documentation.

Technologies: OpenAI GPT-4 API, spaCy NLP, Python (RegEx, os, re, glob, dotenv), Named Entity Recognition (NER), Text Classification, API Integration, Data Cleaning & Chunking, Clinical Documentation Automation, Secure API Handling, Batch Processing, Machine Learning (ML).

Client: Ericsson Dec 2016 – Oct 2022

Location: Stockholm, Sweden

Strategic Leadership and Expertise: Conversational Design, Integration, and Automation

Cross-Functional Leadership & Coordination

Led collaboration across teams in the SDLC, ensuring smooth project execution and stakeholder alignment.

Facilitated communication between developers, designers, and QA teams to meet timelines and resolve roadblocks.

Provided troubleshooting, analysis, and strategic insights for project success.

Supported User Acceptance Testing (UAT) and production deployment.

End-to-End Chatbot Development

Designed and deployed chatbots using conversational platforms, guiding the team from concept to deployment.

Leveraged NLP techniques to enhance chatbot interactions, improving user engagement and satisfaction.

Integration & Collaboration

Worked with technical teams to integrate chatbots with APIs and backend systems, ensuring seamless functionality.

Testing & User Experience Enhancement

Conducted rigorous testing to ensure chatbot security, accuracy, and usability.

Enhanced UX with carousels and menus for intuitive navigation.

Performance Monitoring & Continuous Improvement

Tracked KPIs such as chatbot usage rates, response times, and user satisfaction.

Analyzed data trends to continuously improve chatbot performance and user experience.

Robotic Process Automation (RPA) & Workflow Automation

Automated workflows using Power Automate, improving business process efficiency.

Integrated Power Apps with automation workflows for seamless operations.

Designed and tested Power Automate workflows, including triggers, actions, and conditions for optimal performance.

Project 1: STC Chatbot

Role: Senior conversational Engineer Summary:

Developed an Avaamo-powered chatbot for Ericsson's BDGS Area, optimizing operations by reducing irrelevant inquiries and streamlining incident, change, and FAQ logging. Integrated with JIRA via Ericsson API GW for improved ticket management. Enhanced user experience with carousels, feedback mechanisms, and real-time tracking of KPIs. Hosted the chatbot on SharePoint for easy accessibility.

Responsibilities:

Designed and implemented NLP-driven ticket creation to minimize unnecessary submissions.

Developed use cases for incident creation, change management, and ticket status tracking.

Integrated chatbot with JIRA via Ericsson API Gateway for seamless ticket management.

Improved user experience with carousels, menus, and a feedback mechanism for better interaction.

Enabled request modifications before submission to enhance accuracy.

Monitored KPIs, including usage rates and response times, for continuous optimization.

Provided Logs & Reports for insights and set up alerts for timely issue resolution.

Deployed the chatbot on SharePoint for easy organization-wide access.

Technologies:

API Integration, UX Design, KPI Tracking, Logging & Reporting, JIRA, Ericsson API Gateway, SharePoint, Avaamo, JavaScript

Project 2: SDU Romania PoC Role: Conversational Designer Summary:

Developed a Kore.ai XO Platform chatbot to assist MSDP Area field engineers with incident and alarm tracking by integrating with the incident management system. Implemented NLP and Intent Recognition to provide real-time updates on incident status, priority, and resolution, enabling quick and efficient information access.

Key Responsibilities:

Designed and implemented chatbot workflows for incident tracking, change requests, and alarm monitoring using Kore.ai XO Platform.

Integrated the chatbot with the incident management system through RESTful APIs for real-time data retrieval.

Developed the chatbot’s logic and automation using JavaScript to enhance user interactions.

Implemented NLP-based intent recognition for accurate query interpretation.

Built a real-time update system in Node.js to ensure up-to-the-minute incident tracking.

Utilized GitHub for version control, managing chatbot code and configurations.

Configured Jenkins pipelines to automate testing and deployment processes.

Enhanced API security to ensure safe and efficient data exchange between the chatbot and backend systems.

Technologies Used:

Kore.ai XO Platform, JavaScript, Node.js, RESTful APIs, Intent Recognition, GitHub, Jenkins, API Integration, Real-Time Updates, System Security

Project 3: MELA S&R Chatbot Role: Chatbot Developer Summary:

Developed a chatbot system for handling Customer Service Requests (CSRs) by integrating it with the CSR platform via APIs. Enabled users to create new requests, check status, and modify details before final submission, ensuring accuracy and a seamless experience.

Responsibilities:

Designed and integrated the chatbot with the CSR system for request creation and tracking.

Automated user data gathering (name, contact details, issue description) to streamline CSR issuance.

Implemented API calls for real-time CSR status updates and request modifications.

Ensured a user-friendly experience, allowing users to review and refine details before submission.

Technologies:

Chatbot Development, API Integration, CSR System Integration, User Data Handling, Error Handling, UX Optimization

Project 4: AI Chatbot using Dialogflow CX Role: Dialogflow CX Engineer

Summary: Designed and deployed a Google Dialogflow CX chatbot to automate user interactions, streamline customer service, and enhance engagement. The chatbot leverages state-based conversation design, intent handling, and webhook integrations to provide dynamic responses based on real-time data.

Implemented Google Cloud Platform (GCP) services to ensure scalability and performance, enhancing user experience with efficient API calls and cloud logging.

Responsibilities:

Developed and deployed a state-based chatbot using Google Dialogflow CX for automating user interactions, reducing manual support workload, and improving engagement.

Designed state-based conversation flows, managing intent handling, entity extraction, session persistence, and contextual transitions to ensure smooth interactions.

Implemented webhook integration using Python Flask webhooks hosted on Google Cloud Functions and Cloud Run, enabling dynamic responses based on real-time data.

Integrated Google Cloud Logging for monitoring chatbot interactions, tracking user engagement, and optimizing performance based on insights.

Leveraged Google Cloud Platform (GCP) services to deploy and scale chatbot operations, ensuring high availability and efficient API calls for external integrations.

Developed RESTful APIs to connect the chatbot with CRM and knowledge bases, allowing real- time data retrieval and enhancing conversational accuracy.

Designed the chatbot’s fallback mechanism and error handling workflows to improve resilience against incorrect or unexpected inputs.

Conducted A/B testing to refine conversation flows and improve chatbot containment. Enhanced automated responses for better query resolution.

Technologies: Dialogflow CX, Google Cloud Platform (GCP), Google Cloud Functions, Cloud Run, Python Flask, Google Cloud Logging, RESTful APIs, State-based Conversation Design, Intent Recognition, Entity Extraction, Session Management, Webhook Integration.

Project 5: Blue Prism Role: Developer Summary:

Managed the operational efficiency of a Blue Prism bot, ensuring seamless task processing, real-time monitoring, and performance optimization. Implemented alerts for downtime and processing thresholds, analyzed key performance metrics, and resolved bot failures through session log analysis.

Responsibilities:

Monitored task processing efficiency, tracking average processing times and setting up alerts for delays.

Ensured bot availability with real-time monitoring and downtime notifications.

Logged and analyzed errors, warnings, and performance metrics to improve efficiency.

Oversaw live bots, resource PCs, and control rooms for optimized resource allocation.

Investigated and resolved bot failures through detailed session log analysis.

Technologies:

Blue Prism, Bot Monitoring, Alert Configuration, Performance Metrics Analysis, Error Logging, Root Cause Analysis, Process Optimization

Project 6: One Portal Program Role: Developer

Summary:

Developed and optimized My Support Services workflows using BMC Remedy. Implemented SAP, MWP, and PUC services, created service definitions, and built workflows for BMC Service Broker. Integrated services into BMC ITSM, configured approvals, notifications, and conducted performance monitoring for continuous improvements.

Responsibilities:

Designed and implemented workflows using BMC Remedy for critical services.

Created service definitions, templates, and integrations for BMC Service Broker.

Conducted rigorous testing to ensure workflow efficiency and seamless automation.

Integrated services with BMC ITSM, configuring approval workflows and notifications.

Monitored service performance using analytics and reports to drive improvements.

Technologies:

BMC Remedy, BMC ITSM, Service Workflow Automation, API Integration, Service Performance Monitoring, Testing & Optimization

EDUCATIONAL QUALIFICATIONS:

Bachelor of Technology in Information Technology (2013-2017) - CGPA: 8.2 GITAM University, Visakhapatnam, INDIA

Contact this candidate