
Aziz Bhavanagarwala
Verified Expert in Engineering
Full-stack Developer
Ahmedabad, India
Toptal member since September 8, 2026
Aziz is a full-stack developer with 4 years of experience building production-ready applications with React, Next.js, Node.js, and TypeScript. His expertise spans generative AI (GenAI), RAG, agentic AI, and workflow automation, enabling him to build intelligent, scalable products end to end. Aziz brings strong problem-solving skills, client communication, and a pragmatic approach to delivering high-quality software.
Portfolio
Experience
- Full-stack - 4 years
- React - 4 years
- Web Development - 4 years
- Next.js - 3 years
- RAG Systems - 2 years
- Amazon Web Services (AWS) - 2 years
- Generative Artificial Intelligence (GenAI) - 2 years
- Agentic AI - 2 years
Work Experience
Software Engineer
TOPS Infosolutions Pvt
- Developed a multi-tenant SaaS platform for electric utilities to predict storm-driven grid damage, outages, and restoration resource requirements.
- Built software capabilities that help utility companies anticipate storm impacts and make data-driven decisions around outage response and resource planning.
- Contributed to the development of a scalable SaaS solution focused on predictive analytics and operational planning for electric utility companies.
Founder & Lead Engineer
Schedulio AI
- Designed and built an AI receptionist platform from the ground up, automating customer conversations and appointment booking for clinics across web chat, WhatsApp, and live phone calls.
- Architected a real-time voice pipeline integrating speech-to-text, LLM reasoning, and text-to-speech over Twilio telephony, tuned for low enough round-trip latency to sustain natural conversation.
- Implemented agent orchestration that routes each conversation through intent detection, qualification, availability lookup, booking, and follow-up as discrete stages rather than a single prompt.
- Built multi-tenant architecture with per-clinic data isolation, configurable agent behavior, and HIPAA-conscious handling of patient information.
- Handled the full stack independently: system architecture, back-end services, third-party integrations, the front end, and production deployment on AWS.
- Owned product decisions alongside engineering, defining feature scope and agent behavior based on how clinics actually handle inbound calls.
Software Engineer
Creole Studios
- Delivered end-to-end full-stack applications using React, Next.js, TypeScript, Node.js, and Nest.js, from requirements through production deployment.
- Implemented agentic AI with multimodal reasoning, planning, and tool execution, enabling autonomous support workflows and reducing manual effort by approximately 80%.
- Enhanced React and Next.js applications by optimizing TypeScript code and reducing bundle size, improving page load times by 30% and returning users by 20%.
- Built trainable AI digital workers using files, text, and website data with embeddable chat and voice interfaces for autonomous task execution.
- Automated repetitive business workflows using n8n, Make, and Relevance, eliminating manual operations and saving approximately 12 hours of team effort per week.
- Implemented visual similarity search using FashionCLIP and Pinecone to detect duplicate products with embeddings, confidence thresholds, and metadata filtering.
- Managed direct client communication across projects, gathering requirements, presenting progress, and translating business needs into technical solutions.
Subject Matter Expert
Chegg
- Delivered step-by-step technical solutions to computer science students across data structures, algorithms, operating systems, computer networks, and database systems, maintaining a solution approval rate above 85%.
- Delivered clear, accurate, and high-quality solutions within strict deadlines while maintaining academic integrity.
- Reviewed and debugged student-submitted code across multiple languages, identifying root causes and explaining corrections rather than supplying answers.
- Maintained subject matter expert status in computer engineering through ongoing platform quality reviews and accuracy audits.
Project History
Schedulio.ai | AI Receptionist and Appointment Automation
https://schedulio.aiI architected the real-time voice pipeline speech-to-text, LLM reasoning, and text-to-speech streamed over Twilio telephony - and solved the hard parts of live conversation: keeping round-trip latency low enough to feel natural, handling caller interruptions mid-response, recovering from unclear or partial speech, and maintaining conversation context across a call. Agent orchestration is built with LangGraph, routing each conversation through intent detection, qualification, availability lookup, booking, and follow-up as discrete steps rather than a single prompt. The agent takes real actions via calendar APIs rather than only answering questions.
The platform is multi-tenant, with per-clinic data isolation, configurable agent behavior, and HIPAA-conscious handling of patient information. Built on Next.js, Node.js, PostgreSQL, and Redis, deployed on AWS.
I own the full stack: architecture, back-end services, third-party integrations, the front end, and production deployment.
TradeDesk | Import/Export Management Platform
https://tradedesk.aurexauto.comThe platform is designed to replace fragmented spreadsheet-based processes with a centralized system for managing international trade operations.
AI Digital Workers Platform | Embeddable Chat & Voice Agents
Each agent is trained on customer-supplied material - uploaded documents, raw text, and crawled website content. I built the ingestion pipeline that processes these sources, generates embeddings, and stores them in a vector database, so that every agent answers from its own isolated knowledge base rather than from the general model's knowledge.
The retrieval layer is built with LangChain, pulling relevant context at answer time and passing it to the LLM so responses stay grounded in the customer's material. Agents support both text chat and voice, with speech transcription handled through a real-time transcription service.
I also built the multi-tenant account system, the embeddable widget that customers drop into their own sites, subscription billing, and Slack integration for escalating conversations to a human.
Stack: Next.js, LangChain, OpenAI, Pinecone, Deepgram, MongoDB, Stripe
Service Marketplace Platform | Availability-based Matching and Booking
The central problem was matching. Rather than presenting an unfiltered directory, the system evaluates available professionals against the requirements of each request - schedule availability for the requested time, proximity to the service location, pricing, and the specific services offered - and surfaces the most suitable options.
Booking runs through a real-time acceptance workflow. Requests are offered to matched professionals, and the first acceptance confirms the booking and updates both parties immediately, keeping the calendar consistent and preventing double-booking when requests are concurrent.
I built the matching logic, the availability and scheduling model, the real-time acceptance flow, and the interfaces used by both sides of the marketplace.
Stack: React, Node.js, Supabase
Document Intelligence Platform | RAG-Powered Knowledge Retrieval
I built the system end-to-end. The ingestion pipeline parses PDF and DOCX documents, splits them into chunks aligned with the document structure, generates embeddings, and stores them in a vector database with metadata such as document type, section, and date.
On the query side, the system combines semantic similarity search with keyword matching to retrieve the most relevant passages, then passes them to an LLM to generate a grounded answer. Every response cites the source documents it was drawn from, so users can verify where the information came from.
I also built the web interface for uploading documents, running queries, and reviewing results, along with the APIs connecting it to the retrieval layer.
Stack: Node.js, TypeScript, PostgreSQL with pgvector, OpenAI embeddings, Next.js, AWS.
Automated Tax Filing Platform | Document Data Extraction and PDF Form Generation
The system ingests client data from Excel sheets and extracts complete business profiles - personal details, business and registration information, board of directors, members, and other statutory details - then validates and stores them in a structured database as a single source of truth for each company.
When a filing is due, the user selects the company and the form they need to prepare. The system maps the stored client data to the correct fields in the official fillable PDF and generates a completed document ready for review, rather than having the accountant fill each form manually.
I built the platform end-to-end: the Excel ingestion and data extraction pipeline, the database schema modeling companies and their associated entities, the field-mapping layer between stored records and PDF form fields, the PDF generation service, and the web interface for managing clients and generating filings.
Stack: Node.js, TypeScript, React, Next.js, PostgreSQL, AWS.
Amazon Seller Analytics Platform | MCP-powered AI Agent Over Seller Central Data
The platform ingests reports from the Amazon SP-API and Advertising API on a scheduled basis and stores them in a structured database covering catalog, inventory, orders and sales, and advertising campaigns. On top of the raw data, an insights layer uses LLMs to analyze performance and generate actionable findings, flagging campaigns running above target ACoS, identifying products at risk due to low inventory coverage, and surfacing trends that a seller would otherwise have to find manually across multiple reports.
The core of the product is a conversational AI agent built on an MCP server exposing 80+ purpose-built tools over the seller's data. Instead of navigating dashboards, users ask questions directly, such as how a specific campaign is performing, last month's sales for a given product, inventory, or fee breakdowns, and the agent selects and calls the appropriate tools to answer from their own live data.
Education
Bachelor's Degree in Computer Engineering
Marwadi University - Rajkot, India
Certifications
Certificate of Completion: Al Fluency Framework & Foundations
Anthropic
Next.js: Creating and Hosting a Full-stack Site
Database Programming with SQL
Oracle
Data Structures and Algorithm
Coursera
Skills
Libraries/APIs
React, Node.js, REST APIs, API Development, Stripe, Stripe Connect, Twilio API, SendGrid API, Claude API, WebRTC, Google Calendar API, Stripe API, Playwright, React Testing Library, Puppeteer
Tools
Claude Code, Claude, GitHub, Git, Shadcn, Google Analytics, MongoDB Atlas, GraphRAG, Codex, Drizzle ORM, Claude Agent SDK, n8n, TanStack Router, Vitest, Webpack, Expo, Prisma, Jira
Languages
TypeScript, JavaScript, CSS, SQL, HTML, CSS3, HTML5, GraphQL, Python
Frameworks
Next.js, Express.js, Redux, Tailwind CSS, NestJS, Fastify, LlamaIndex, Agentic Frameworks, Serverless Framework, Jest, React Native, Payload CMS, LangGraph
Paradigms
Database Design, Web Application Architecture, Clean Code, Search Engine Optimization (SEO), Model Context Protocol (MCP), Microservices, Automation, Role-based Access Control (RBAC), HIPAA Compliance, Responsive Web Design (RWD), DevOps, Microfrontends, Agile Software Development, Testing
Platforms
Amazon Web Services (AWS), Vercel, AWS Lambda, Twilio, Firebase, Replit, Google Cloud Platform (GCP), Webflow, Clerk, Telnyx, LiveKit, Docker
Storage
Databases, PostgreSQL, MySQL, Redis, Google Cloud, MongoDB, Data Pipelines
Industry Expertise
Healthcare
Other
Agentic AI, Full-stack, Web Development, Full-stack Development, API Integration, Landing Pages, Agentic Coding, Front-end, Back-end, Vendor Management, RESTFul APIs, RAG Systems, Team Collaboration, Generative Artificial Intelligence (GenAI), Data Structures, Supabase, OpenAI APIs, Retrieval-augmented Generation (RAG), Large Language Models (LLMs), LangChain, AI Agents, AI Integration, APIs, Security, Artificial Intelligence (AI), AI-generated Code, AI-assisted Development, Technical Leadership, Google Analytics 4 (GA4), Web Marketing, AI Automation, Server-side Rendering (SSR), OpenAI SDK, User Experience (UX), AI Tools, SaaS, Core Web Vitals, SEO Tools, eCommerce, Workflow Automation, Workflow Automation & System Integration, Supabase Auth, Web Applications, Cursor AI, Optimization, Performance, Third-party APIs, Debugging, Auditing, Progressive Web Applications (PWAs), Software as a Service (SaaS), Data Privacy, Architecture, Text-to-Speech (TTS), Speech-to-Text (STT), Web Scraping, LLM Integration, Workflows, User Interface (UI), Design, Solution Architecture, System Architecture, Prompt Engineering, Agentic AI Systems, Vector Databases, Minimum Viable Product (MVP), Gemini API, Code Review, Healthcare Software, WebSockets, AI Development, Streaming, LLM Agents, Audits, Scalability, Legal, Knowledge Graphs, WCAG 2, Search Engines, Accessibility, System Integration, Authentication, Authorization, HIPAA, Webhooks, Calendars, Chromium, Railway, OpenAI, Dashboards, Data Visualization, Integrations, Back-end Development, Front-end Development, Software Development, Software Development Lifecycle (SDLC), Business Services, ElevenLabs Solutions, AI Voice Agents, Conversational AI, Fine-tuning, Intent Classification, OpenRouter, Real-time Audio Processing, Application Security, Cloud Infrastructure, CI/CD Pipelines, Code Auditing, GPS, Multi-tenant SaaS, Pinecone, ChromaDB, Vite, Software Architecture, Affiliate Marketing, Zustand, Resend, CMS, Amazon DocumentDB, AI Architecture, AI Engineering, A/B Testing, Analytics, Logistics, Shipping, CRM, AI Agent Orchestration, Semantic Search, Embeddings from Language Models (ELMo), Data Extraction, Document Automation, Server-side PDF Generation, Data Analysis, Problem Solving, Scheduling Tools, Geolocation
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
Share your needs
Choose your talent
Start your risk-free talent trial
Top talent is in high demand.
Start hiring