Shalini H.

Key Skills

Software

Other

Top Subject Matter

Travel/aviation (flight search and preferences)

Internal knowledge discovery (document-grounded QA)

Travel planning (multi-agent recommendation)

Top Data Types

Text

Top Task Types

Function Calling

Question Answering

Text Generation

Text Summarization

Prompt + Response Writing (SFT)

Freelancer Overview

Software Engineer delivering AI-powered product features such as Natural Language Filter. Brings 7+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include C#, .NET, JavaScript, Python, Azure OpenAI, Azure AI Foundry, and Other. Demonstrated critical analysis, attention to detail and strong problem solving skills. Education includes Master of Information Systems, University of Melbourne (2022) and Bachelor of Technology, Manipal University (2017). AI-training focus includes data types such as Text, Computer Code, and Programming and labeling workflows including Function Calling, Question Answering, and Text Generation.

Labeling Experience

Webjet — Software Engineer (AI-powered Natural Language Filter)

TextFunction Calling

Built an AI-powered Natural Language Filter that converts users’ plain-language travel preferences into structured filter intent for flight search. Designed prompt/service contracts to produce structured LLM outputs and reduce ambiguity for safer downstream integration. Implemented observability pipelines to track AI feature performance using user inputs, model clarifications, and confidence signals. • Developed and integrated structured intent generation for flight search workflows. • Created prompt contracts and structured outputs for reliable machine use. • Instrumented request volume, failures, and analytics for iterative improvement. • Diagnosed production issues using structured logs and model-related signals.

2022 - Present

Selected AI Projects — Workflow Automation (n8n/Make/Manus)

OtherPrompt + Response Writing (SFT)

Created AI-assisted automation workflows to support content generation pipelines and CRM lead capture. Implemented workflow logic that combines generative steps with automation tools for small-business use cases. Focused on integrating AI-generated content into operational processes via repeatable workflow runs. • Built AI-assisted content generation workflow steps. • Automated CRM lead capture using workflow orchestration. • Integrated multiple automation tools to support end-to-end generation pipelines.

2025 - 2026

Selected AI Projects — Azure AI Search & Foundry Knowledge Agent

TextQuestion Answering

Built a RAG-based knowledge agent that answers questions using retrieved internal documentation as grounded source material. Used Azure AI Foundry and Azure AI Search to index knowledge and support response generation grounded in curated documents. Focused on improving knowledge discovery through retrieval and evaluation of agent outputs. • Created RAG pipeline for question answering over indexed documentation. • Implemented grounding to ensure answers are based on curated sources. • Used Azure AI tools to orchestrate retrieval and response generation.

2025 - 2026

Selected AI Projects — AWS Bedrock Fare Rules Summarization

OtherTextText Summarization

Designed a system to translate complex airline fare rules into customer-friendly summaries. Used Amazon Bedrock and the AWS Strands Agent SDK to generate readable explanations from structured or semi-structured fare rule inputs. Produced a hackathon-ready solution focused on clarity, usability, and accurate summarization. • Built fare-rule summarization into customer-friendly text. • Leveraged Bedrock and agent SDK for automated interpretation and rewriting. • Ensured outputs were intended for end-customer consumption.

2025 - 2025

Selected AI Projects — Microsoft AutoGen Multi-Agent Travel Planner

OtherTextText Generation

Developed a multi-agent AI travel planner where users describe holiday preferences in natural language and receive personalized travel and hotel recommendations. Implemented multi-agent orchestration to transform preference descriptions into structured recommendations. Delivered the solution as part of a time-boxed hackathon project, emphasizing rapid prototyping and user-facing outputs. • Orchestrated multi-agent flows to generate travel and hotel recommendations. • Converted natural-language preferences into actionable itinerary outputs. • Built a hackathon prototype demonstrating end-to-end agent behavior.

2025 - 2025

Education

Work History