I help startups and product teams build, stabilize, and scale AI-powered SaaS platforms.
I am a Senior Full-Stack / AI SaaS Engineer with 11+ years of experience across TypeScript, React, Next.js, Node.js, PostgreSQL, Supabase, Redis, Go, cloud infrastructure, and LLM-powered workflows. I work best in small, high-ownership teams where I can take ambiguous requirements, understand the existing system, and ship production-ready features without breaking what already works.
My recent experience includes AI assistants, realtime systems, multi-tenant SaaS platforms, Stripe billing and entitlements, class scheduling APIs, secure file access, webhooks, browser automation, backend integrations, and production debugging.
At MSBAI, I helped build AI products for scientific and engineering workflows, including an AI assistant for CFD simulation planning, a mining-permit drafting assistant, and an HPC/supercomputer resource-management platform. I worked on LLM workflows, retrieval systems, realtime desktop sharing, resumable job pipelines, and admin tooling.
At LSAT Demon, I worked on production TypeScript systems for class scheduling, student registration, plan-based access control, recurring class behavior, downgrade cleanup, Zoom workflows, and regression-safe service logic.
At Handlet, I work on an AI-native SaaS platform using Next.js, Supabase, Stripe, realtime inbox workflows, Unipile integrations, secure attachment access, entitlement enforcement, RLS, webhooks, migrations, and CI/CD.
I bring strong full-stack execution, backend architecture, product sense, testing discipline, and the ability to move fast without creating long-term technical debt.
At LSAT Demon: Designed and implemented a workspace-centric SaaS foundation with RBAC, invitations, Stripe billing, e...
At LSAT Demon: Designed and implemented a workspace-centric SaaS foundation with RBAC, invitations, Stripe billing, entitlement enforcement, and feature-gated workflows, creating a reusable platform layer for B2B AI product development serving 50K users
- Architected AI workflow infrastructure for persistent memory, retrieval, and tool-state orchestration, enabling multi-step AI workflows to resume with consistent context across sessions.
At Handlet: Worked on a multi-tenant AI SaaS platform using Next.js, Supabase, Stripe, realtime inbox flows, Unipile integrations, and secure attachment access.
- Implemented and reviewed entitlement enforcement, billing flows, webhooks, RLS behavior, CI/CD workflows, and Supabase migration discipline.
- Used Playwright-style E2E validation and TypeScript tests to protect critical user flows.
At MSBAI: Developed the admin platform for a large language model (LLM)-powered assistant that diagnoses and resolves technical issues in scientific workflows.
- Debugged AI-generated solutions, optimized HPC job pipelines in real time.
- Supported users in adapting LLM outputs into production-ready code and tools.
- Troubleshot system-level issues (e.g., job schedulers, CLI behavior) and provided clear technical guidance.
At Xintra & Scholarly:
- Created a SaaS starter template reused across multiple products: a live test-taking platform, an AI-based 3D modeling asset marketplace, and a 1:1 mentoring app.
- Built a real-time desktop sharing tool (15+ FPS, low latency) enabling remote simulations and training.
- Developed a no-code app builder for rapid prototyping of internal apps.
- Built a Gmail/Outlook add-on to automate transport bid margin calculations from emails.
Stack: Typescript (React, Next, Node), Rust (Axum, Dioxus, Tauri), DB (Supabase, PostgreSQL, MongoDB), cloud
Built apps that aggregated data from multiple source and shared it in stakeholder’s preferred format or process
- HR Das...
Built apps that aggregated data from multiple source and shared it in stakeholder’s preferred format or process
- HR Dashboard: HR can track various insurance policies and claims of their company’s employees
- Company Dashboard: Companies can purchase customised policies from an aggregated list of insurers; They can maintain a running balance to cover the changes in their no of employee at the policies; Compare and receive bids for their customised policies across the aggregated list of insurers
- Employee Dashboard: Employee can purchase customised policies; add/remove dependants and track their claims
- Standardised development stack to use Firebase functions and Next; with logging-monitoring-tracing and testing
Stack: Next, React, Styled Components, Tailwind CSS, GraphQL, NodeJs, Cypress, Github CI/CD
As a part of engineering effort at nfer Clinical nSights, i worked on apps like:
- AI Studio: Users can upload ...
As a part of engineering effort at nfer Clinical nSights, i worked on apps like:
- AI Studio: Users can upload and tag dataset. Users run models against these datasets. Users can review the model's prediction and feed this back to the model. workspaces: Svelte UI to track uptime and cost of GCE instances that run models in AI studio; CRUD
- Triangulation: also known as drug lifecycle management, rate targets after correlation between triangulation of literature, molecular and real world evidence.
- Real-World Evidence: Automated real-time synthesis of key opinion and clinical activity of biomedical concepts
Develop and maintain libs that work across nference's suite of apps like:
- core-ui: UI component library with storybook and nference's design language
- annotator: user-generated spatial annotations over non-text data like pathology images, ECG waveforms and maps
- Notification service: middleware to execute Async jobs and caches results for rate limiting requests, book-keeping, notification in the backend and a collection of web socket based subscription hooks to consume real time data
Stack: React, Tailwind CSS, GraphQL, Golang, NodeJs