I’m an AI Software Engineer focused on building end-to-end AI systems that actually work in the real world. My strength lies in combining prompt engineering, model understanding, and solid software engineering to turn ideas into production-ready AI products—not just demos. I’ve built multiple AI-driven applications across domains like legal-tech, therapy chatbots, regulatory intelligence, automation tools, and conversational agents. My work includes scraping and processing large-scale unstructured data (PDFs and dynamic websites), LLM-based summarization, retrieval-augmented generation (RAG), vector databases, and human-in-the-loop AI systems. I’ve worked hands-on with a wide range of LLMs including Gemini, OpenAI, Mistral, Grok, DeepSeek, and several Hugging Face models, and I understand which models perform best for specific use cases. Beyond prompting, I also handle the technical depth—backend architecture, databases, APIs, deployment, and scalability. Before moving fully into AI, I worked as a Java Developer at Accenture, which gave me a strong engineering foundation. Today, I focus on building modular, explainable, and user-centric AI systems, with a strong interest in automation, AI agents, and human-AI collaboration.

Chaitanya Bandawat

I’m an AI Software Engineer focused on building end-to-end AI systems that actually work in the real world. My strength lies in combining prompt engineering, model understanding, and solid software engineering to turn ideas into production-ready AI products—not just demos. I’ve built multiple AI-driven applications across domains like legal-tech, therapy chatbots, regulatory intelligence, automation tools, and conversational agents. My work includes scraping and processing large-scale unstructured data (PDFs and dynamic websites), LLM-based summarization, retrieval-augmented generation (RAG), vector databases, and human-in-the-loop AI systems. I’ve worked hands-on with a wide range of LLMs including Gemini, OpenAI, Mistral, Grok, DeepSeek, and several Hugging Face models, and I understand which models perform best for specific use cases. Beyond prompting, I also handle the technical depth—backend architecture, databases, APIs, deployment, and scalability. Before moving fully into AI, I worked as a Java Developer at Accenture, which gave me a strong engineering foundation. Today, I focus on building modular, explainable, and user-centric AI systems, with a strong interest in automation, AI agents, and human-AI collaboration.

Available to hire

I’m an AI Software Engineer focused on building end-to-end AI systems that actually work in the real world. My strength lies in combining prompt engineering, model understanding, and solid software engineering to turn ideas into production-ready AI products—not just demos.

I’ve built multiple AI-driven applications across domains like legal-tech, therapy chatbots, regulatory intelligence, automation tools, and conversational agents. My work includes scraping and processing large-scale unstructured data (PDFs and dynamic websites), LLM-based summarization, retrieval-augmented generation (RAG), vector databases, and human-in-the-loop AI systems.

I’ve worked hands-on with a wide range of LLMs including Gemini, OpenAI, Mistral, Grok, DeepSeek, and several Hugging Face models, and I understand which models perform best for specific use cases. Beyond prompting, I also handle the technical depth—backend architecture, databases, APIs, deployment, and scalability.

Before moving fully into AI, I worked as a Java Developer at Accenture, which gave me a strong engineering foundation. Today, I focus on building modular, explainable, and user-centric AI systems, with a strong interest in automation, AI agents, and human-AI collaboration.

See more

Experience Level

Expert
Expert
Expert
Expert
Expert
Expert

Work Experience

Freelance AI Software Engineer at Upwork
October 1, 2024 - Present
Delivered end-to-end applied AI and ML solutions for clients in legal, finance, and customer service, integrating LLMs, NLP, and scalable data pipelines. Built AI-powered document processing pipelines handling 10,000+ documents with Playwright and Scrapy; reduced manual review by 70%. Developed retrieval-augmented generation (RAG) systems with GPT-4, Gemini Flash, LangChain, and Pinecone, achieving 92% retrieval precision. Designed and deployed AI chatbots and summarization systems handling 2,000+ queries/month with sub-2s latency. Implemented PostgreSQL + FastAPI backends for secure metadata storage and query tracking; automated compliance reporting reducing turnaround from 3 days to 4 hours.
NLP Engineer at Welsh Revenue Authority
June 1, 2023 - September 1, 2023
Collaborated with government stakeholders to automate tax query guidance using NLP and semantic search. Developed a Flask-based NLP API for tax guidance retrieval, leveraging embeddings + Pinecone. Processed and embedded 20,000+ tax-related documents for faster retrieval. Improved retrieval accuracy by 35% over keyword search via transformer-based embeddings. Reduced tax query resolution time by 60% with a chatbot delivering relevant guidance directly to users.
Application Development Analyst at Accenture
December 1, 2021 - August 31, 2022
Led insurance domain platform modernization, integrating multiple systems into a single application, reducing onboarding workflows by 40%. Optimized Spring Boot + Angular app, reducing load time from 10s to 3s and improving customer retention. Cut incident resolution time by 30% by refactoring legacy code and implementing structured documentation. Delivered a critical project 3 weeks ahead of schedule by applying Agile methodologies.
Application Development Associate at Accenture
August 1, 2020 - November 30, 2021
Promoted 9 months early (top 5% performers) for strong project delivery.

Education

MSc at Cardiff University
September 1, 2022 - September 1, 2023
Bachelor of Technology at Banasthali Vidyapith
July 1, 2016 - July 1, 2020

Qualifications

Add your qualifications or awards here.

Industry Experience

Software & Internet, Government, Professional Services, Financial Services
    paper Anita Therapy ChatBot

    Anita is a compassionate AI-powered therapy chatbot built using BERT-based intent classification, semantic search with FAISS, and LLM-driven responses via Gemini Pro. It offers emotionally aware conversations and generates speech output using Sarvam AI’s multilingual TTS API.

    You can check the code and the demo here: https://www.twine.net/signin