Purvangi Gor

Data Scientist

Ahmedabad, Gujarat, India1 yr 8 mos experience

AI EnabledAI ML Practitioner

Key Highlights

Expert in developing production-grade AI solutions.
Hands-on experience with generative AI and NLP.
Strong focus on making AI practical and impactful.

Stackforce AI infers this person is a SaaS-focused AI Engineer with expertise in NLP and generative AI solutions.

Contact

Skills

Core Skills

Natural Language Processing (nlp)Generative AiData ScienceLarge Language Models (llm)

Other Skills

Python (Programming Language)Google Add-onsContextual inferenceRule-based modelsHTMLCSSJavaScriptFlaskNatural Language UnderstandingVideo ProcessingMetadata ManagementMachine Learning AlgorithmsQdrantChatbot DevelopmentChatbot Responses

About

I build AI systems that move from concept to production and I write about what I learn along the way. With hands-on experience across generative AI, natural language processing, voice intelligence, and intelligent automation, I have delivered production-grade AI solutions including domain-specific AI assistants, health-focused recommendation engines, NLP-powered video search systems, and multilingual TTS pipelines integrated into platforms like Gmail and Microsoft Teams. My focus areas: GenAI, Agentic AI, LLMs, NLP, and intelligent system design. What I bring to the table: the ability to take an AI concept from proof-of-concept through production deployment, combined with a deep interest in making AI practical, accessible, and impactful for real-world use cases. I share my insights on AI tools, trends, and technical deep-dives through my writing here on LinkedIn. Open to: speaking opportunities, AI tool reviews, collaborations, and conversations about where AI is heading next.

Experience

1 yr 8 mos

Total Experience

1 yr 8 mos

Average Tenure

1 yr 8 mos

Current Experience

Cipio.ai

2 roles

Data Scientist

Sep 2024 – Dec 2024 · 3 mos · Ahmedabad, Gujarat, India · On-site

Spearheaded the development of a long-form video processing system, enabling users to input natural language queries to retrieve contextually relevant short clips from lengthy videos. Integrated features such as clip trimming, adjustable timestamp selection, transcript viewing, and direct video download for enhanced usability.
Designed and implemented a natural language search system for a large video clip database. Created and stored metadata to enable users to search based on context (e.g., gender, location, remembered phrases), which returned structured, relevant clips via our Video Fusion product.
Conducted continuous proof-of-concept (PoC) testing with various Large Language Models (LLMs) to identify the most effective solution for video content understanding and semantic search.
Worked on segmentation model fine-tuning to improve accuracy and adaptability for specific video datasets.
Developed customizable video captioning pipelines and integrated text-to-speech (TTS) systems with multilingual voice selection, boosting personalization capabilities for marketing content creation.
Managed infrastructure tasks such as database connections, SQL scripting, cron job scheduling, virtual machines, and cloud services, ensuring smooth deployment and production integration.
Leveraged ChatGPT and other AI tools to simplify and accelerate complex development tasks, contributing to rapid prototyping and experimentation.

Python (Programming Language)Natural Language Processing (NLP)Generative AI

Data Science Intern

Feb 2024 – Aug 2024 · 6 mos · Ahmedabad, Gujarat, India · On-site

Developed and optimized algorithms for video segment categorisation, achieving 85% accuracy in categorizing video content as part of the VideoFusion project.
Implemented a semantic search using the Qdrant vector database to enhance the retrieval of relevant video segments through text and image queries.
Automated key processes including transcript generation, key frame extraction and emotion detection, significantly improving workflow efficiency and consistency.
Collaborated with the data science team to deploy machine learning models for image recognition, content classification, and text detection, ensuring reliable and accurate results.
Played a key role in data handling, including data cleaning, analysis, and scripting tasks, ensuring seamless integration and optimal performance in video content creation workflows.
Explored and experimented with new generative models, generating proof of concepts (POCs) to demonstrate their potential applications in enhancing video content creation and marketing strategies.

Large Language Models (LLM)Data Science

Dxfactor

Data Scientist

Sep 2024 – Present · 1 yr 8 mos · Ahmedabad, Gujarat, India · On-site

Built a domain-specific AI employee assistance agent integrated into Gmail via Google Add-ons, utilizing CardService for intelligent UI rendering and conversational query support.
Engineered a health-focused recommendation engine to provide personalized insights based on structured payloads (e.g., lab results, health goals like weight loss), leveraging contextual inference and rule-based models.
Developed and deployed an interactive cloud service analysis dashboard comparing AWS, Azure, and hybrid services using HTML, CSS, JavaScript, and Flask, supporting rich visualizations, tabular formats, and dynamic, user-friendly data interactions.
Designed and exposed robust RESTful APIs to support scalable machine learning solutions and analytics workflows.
Created KPI dashboards for project performance analysis, enabling real-time decision-making and progress tracking.
Integrated Microsoft and Google Single Sign-On (SSO) for seamless and secure authentication across internal platforms and client solutions.
Developed an HR virtual assistant chatbot integrated into Microsoft Teams, capable of answering employee queries regarding company policies using natural language understanding.
Built a voice-personalized executive assistant with Text-to-Speech (TTS) capabilities using cloned voice models for interactive, speech-enabled interfaces.
Spearheaded latency reduction and performance optimization initiatives to enhance frontend responsiveness and backend compute efficiency.
Executed end-to-end development for all solutions—from architecture and prototyping to deployment and client handover.
Delivered technical PoCs for model analysis and experimental validation in support of scalable AI adoption.
Actively managed client communication, aligning technical delivery with evolving business needs and stakeholder feedback.