HI|||||||| THERE
I'M DEEPAKSHI
AI ENGINEERFULL STACKDEVELOPER
Passionate about creating meaningful digital experiences
21
2025 Grad
B.Tech
CSE (AI & DS)
7.8
CGPA
I am an AI Engineer with hands-on experience building and deploying real-world AI systems using LLMs, RAG pipelines, and conversational AI. I specialize in scalable backend systems, real-time voice AI, and multi-agent architectures.
I have shipped production-grade systems involving voice automation, document intelligence, and AI-powered workflows — with a focus on low latency, cost-efficiency, and developer ergonomics.
I enjoy solving real-world problems using AI and building systems that improve efficiency and reduce manual effort.
Personal Details
deepakshis21@gmail.com
Location
Gurugram, India
Freelance
Available
Hobbies & Interests
SKILLS
The toolkit I use to ship Gen-AI products: LLMs, RAG, voice AI, multi-agent orchestration, and the backend systems behind them.
Education
B.Tech in Computer Science (AI & Data Science)
Chandigarh Group of Colleges, Landran, India
2021 - 2025 | CGPA: 7.8/10 (78%)
PROJECTS
A curated selection of my best development projects that showcase my skills and expertise.
AI Voice Calling Assistant
★ Featured• Built a real-time AI voice calling system supporting inbound and outbound automation using a multi-agent architecture.
• Developed a WhatsApp AI agent for conversational interactions across customer touchpoints.
• Built a campaign system to manage large-scale calling workflows with concurrency and billing controls.
• Implemented RAG for context-aware conversations grounded in customer knowledge bases.
• Designed low-latency real-time communication using LiveKit; integrated Twilio + telephony APIs for scalable calling.
• Impact: automated calling workflows end-to-end, significantly reduced manual effort, shipped a production-ready system.
AI Task Manager — Multi-Agent System
★ Featured• Developed a multi-agent AI system to manage tasks via natural language and voice interaction.
• Built specialized agents for task creation, tracking, and completion with hand-off via LangGraph workflows.
• Integrated voice interaction for hands-free task management; enabled real-time updates and live task monitoring.
• Impact: reduced manual task-management effort by ~40%.
Chat with PDF — RAG System
★ Featured• Built an end-to-end RAG-based document Q&A system with PDF parsing, chunking, embeddings, and retrieval.
• Created an embedding pipeline backed by a vector database (ChromaDB) for context-aware answers.
• Designed a full-stack interface for user interaction over uploaded documents.
• Impact: improved document search and retrieval accuracy versus keyword baselines.
LLM Audio QA Automation
• Developed an AI system to process audio files and extract structured insights end-to-end.
• Converted speech to text using Whisper ASR; applied LLMs for question–answer extraction.
• Generated structured reports automatically from raw audio inputs.
• Impact: automated previously manual audio-analysis workflows.
Conversational AI (Dialogflow CX)
• Built a multi-agent conversational AI system with agents for chat, task assignment, and tracking.
• Integrated RAG for memory and long-context grounding across conversations.
• Created structured workflows for user engagement and intent routing.
LLM Cost Optimization
• Optimized LLM performance and reduced inference cost across production workloads.
• Applied prompt optimization, response caching, and model selection / routing.
• Reduced token usage and latency on common request paths.
• Impact: achieved ~50% cost reduction.
Sales Prediction Model
• Built an ML model to predict store sales using historical data.
• Performed data preprocessing and feature engineering on retail time-series data.
• Applied ML algorithms for prediction and evaluated model performance against baselines.
EXPERIENCE
Currently working on multi-agent AI systems, voice AI with real-time interaction, and advanced RAG pipelines.
AI Engineer
sftwtrs.ai
• Built real-time conversational AI systems for inbound and outbound voice automation.
• Designed voice AI pipelines using LiveKit + Twilio with low-latency real-time communication.
• Developed RAG-based systems for context-aware conversations grounded in customer knowledge bases.
• Owned scalable backend systems handling concurrency, billing, and agent workflows.
• Worked across cloud deployment on GCP and AWS for production AI services.
SERVICES
Bringing ideas to life with full-stack development and scalable solutions tailored to your needs.
AI & LLM Solutions
Building production-grade RAG systems, autonomous agents, and fine-tuning LLMs (GPT-4, Llama) for specific enterprise use cases.
Voice AI Systems
Developing real-time, low-latency voice assistants using LiveKit and WebRTC to enable natural, conversational user interactions.
Full-Stack Development
Creating scalable web applications with Next.js, React, and robust backends using FastAPI/Python, tailored to business needs.
Backend & API Design
Designing high-performance RESTful APIs, Microservices, and WebSocket architectures for seamless data communication.
Cloud & DevOps
Deploying secure, scalable applications on AWS/GCP using Docker and CI/CD pipelines to ensure reliability and uptime.
Automation Pipelines
Engineered automated workflows and data extraction pipelines to reduce operational costs and increase efficiency.
CONTACT ME
Have a project in mind? Feel free to reach out and let's create something amazing together.