Sharan Shyamsundar
AI Engineer & Data Scientist
Building production-grade agentic AI systems and full-stack AI products that work in the real world.
About Me
Founding AI Engineer · Full-Stack Builder · Professional Vibe Coder
AI Engineer with 5+ years across AI engineering, data science, and quantitative analytics. Currently at Quid as the first technical hire, where I architect and ship production-grade multi-agent systems that automate enterprise accounting workflows end-to-end.
I specialize in turning LLM capabilities into reliable, scalable products — from RAG pipelines and fine-tuned models to full agentic systems with tool access and memory. Comfortable operating across the entire stack: agent architecture, ERP integrations, FastAPI backends, and React dashboards.
Currently open to exciting opportunities in AI Engineering and Applied AI roles.
Agentic AI Systems
Designing multi-agent architectures with tool access, memory, and RAG for production-grade, domain-specific automation.
LLM Fine-Tuning & RAG
Optimizing LLM pipelines with PEFT, LoRA, and retrieval-augmented generation for accuracy, efficiency, and scalability.
Full-Stack AI Products
Building end-to-end AI products — from agent backends and API layers to interactive frontend dashboards.
Data Science & Analytics
Applying classical ML, financial analytics, and data engineering to drive decision-making across industries.
Experience
- Joined as the first technical hire; evolved early LLM wrapper prototypes into production multi-agent systems with tool access, memory, and RAG — establishing the core AI architecture powering the Quid platform.
- Built Quid AI for Slack — an AI-powered expense management bot enabling users to upload receipts directly in Slack, with automated field extraction, metadata mapping, expense report submission, and real-time analytics reporting.
- Architected a multi-agent accounting system using the Agno framework to automate invoice sync and bookkeeping into DATEV/Lexware, with agents operating as accountants with access to ERP resources, company history, and metadata.
- Contributed across the full product lifecycle including React/TypeScript frontend dashboard components, FastAPI backends, and cloud deployments on AWS and GCP.
- Designed and deployed Eddy, a Retrieval-Augmented Generation (RAG) powered internal helpdesk, reducing query resolution time
- Developed interactive dashboards to streamline supply chain analysis and optimize decision-making
- Delivered multiple proof-of-concept projects showcasing advanced analytics capabilities
- Automated investment workflows and developed intuitive tools for streamlined decision-making
- Led ESG-related financial analytics products ensuring regulatory compliance
- Recruited and trained junior quant analysts, fostering high-performing teams
Education
Tech Stack
Languages
AI & LLMs
Cloud & DevOps
Featured Projects
PEFT Methods Applied to Quantized LLMs
Explored the integration of Quantization and Parameter-Efficient Fine-Tuning (PEFT) methods to optimize LLM finetuning. Achieved substantial improvements in memory efficiency and computational performance for resource-constrained environments.
Report
Foodle - Personal AI Cooking Assistant
An innovative cooking assistant leveraging ASR and LLMs to enhance culinary experiences. Transcribes cooking videos, generates recipes, and offers hands-free voice interactions for real-time cooking assistance.
Multi Agent Article Writer
AI-powered multi-agent framework using CrewAI to automate content creation. Assigns specialized agents for planning, writing, and editing, ensuring high-quality, well-structured articles with enhanced coherence and readability.
ZENIX - Natural Language to XML
Multi-agent AI solution that translates natural language into precise SQL queries while generating necessary XML business objects. Enhanced user experience with intelligent navigation link generation for the JiVs IMP platform.
Pitch Perfect
AI-powered cover letter generator using intelligent prompts and customizable templates. Helps highlight experience and skills for every application, saving time and boosting interview chances for job seekers.
Multi Agent Project Planner
AI-powered project planning app built with CrewAI. Automatically creates clear, step-by-step project plans with tasks and milestones intelligently assigned to team members for organized and efficient project delivery.
Let's Connect
I'm always interested in exploring new opportunities and collaborations. Whether you have a project in mind or just want to connect, feel free to reach out!
Or email me directly at
de.sharan17@gmail.com