Hi, my name is
Abhay Bisht
I build & ship AI-powered products.
GenAI Full-Stack Engineer | LLMs, RAG | Building & Shipping AI-Powered Products End-to-End
Actively seeking full-time opportunities.
see my resumewhat i do
GenAI Full-Stack Development
- Integrating LLMs and Generative AI capabilities (RAG, AI agents, automation) into full-stack applications.
- Building scalable, responsive web applications using React.js and Next.js, integrated with AI-powered features.
- Designing and developing production-ready REST APIs using Node.js, FastAPI, Express.js, TypeScript, and JavaScript.
- Building cross-platform mobile applications with React Native.
- Designing, managing, and optimizing application databases using MongoDB, PostgreSQL, and SQL-based systems.
Python & Data science
- Experience in Exploratory Data Analysis (EDA) and data visualization for extracting insights from complex datasets.
- Experience in data cleaning, preprocessing, and transformation using Pandas and NumPy.
- Experience in web scraping and data extraction using BeautifulSoup and related tools.
- Experience in building interactive data-driven UIs and dashboards using Streamlit.
where i've worked
SDE-1 @Zycus (Tario.ai)
August 2025 - Present
- Engineered a core Agentic AI workflow using LLMs and RAG to generate personalized outreach emails from lead data and knowledge bases, complemented by an interactive chat interface enabling users to edit AI-generated messages and save refined outputs as reusable templates across campaign workflows.
- Built an LLM-powered call analysis dashboard with audio playback and LLM-generated transcripts, featuring a config-driven, widget-based UI with drag-and-drop widgets for tenant-specific customization.
- Developed a Campaign Management UI and mapping agent, enabling users to upload leads, auto-map fields, and configure multi-channel sales sequences (email, call, LinkedIn).
- Developed the complete end-to-end dialing flow, enabling users to create dialing lists, upload CSVs, and perform manual one-to-one as well as parallel dialing using the Twilio Voice SDK.
- Implemented call disposition workflows to capture user feedback and drive lead scoring and prioritization, determining optimal callback times, next actions, and sorting order.
- Built asynchronous call recording processing using SQS, with recordings stored in S3.
- Established tenant-level data isolation across Twilio using parent accounts, sub-accounts, and TwiML configurations, ensuring secure multi-tenancy.
- Integrated Ably for real-time event updates and leveraged SQS and SNS for reliable event processing and real-time notifications, delivering immediate visibility into agent execution and workflow progression.
- Tech Stack - TypeScript, Python, React.js, Redux Toolkit, Storybook, Tailwind CSS, Material UI, Twilio, FastAPI, AWS ( SQS, SNS ), MongoDB, React Query, LangChain, LangGraph, LangSmith, RAG, dnd-kit.
Contributions
featured projects
Zycus (Tario.ai)
An AI-powered SDR platform for modern sales teams.
Built complete sequential & parallel dialing, call analysis dashboard, RAG powered chat interface for outreach messages, Campaign management UI.
314e Corporation (Jeeves)
Jeeves is an AI-powered Just-in-Time training platform that uses microlearning to help EHR users accomplish workflow tasks quickly.
Built screen & web cam recorder, tipsheet asset, interactive & non-interactive annotation editor, asset stages & activity panel.
Gappy.ai
Gappy connects to your apps, understands your workflows, and finishes tasks for you. Send reports, manage projects, crunch numbers, or run multi-step automations—just by asking.
Built complete copilot sdk, dashboard for managing agents, knowledge bases, tools, knowledge graphs, copilot theme.
my projects
what's next
get in touch
I’m currently looking for new opportunities.
Connect with me on Linked In
Connect