HS.
All projectsEngineering case study

DIGITAL ALPHA · AI PLATFORM

epiphAI — agent APIs and AI interfaces

Built agent configuration APIs, organization-scoped access, credit logs, in-chat charts, and streaming recovery.

Himanshu SharmaCase study updated

My role

At Digital Alpha Platforms, I contributed backend and frontend features to epiphAI, an AI-agent and automation platform.

Backend building blocks

I implemented Django REST APIs for agents, configurations, LangFlow management, organization API-key authentication, endpoint permissions, and credit-consumption logs. Across my Digital Alpha work, I improved Django API response times by approximately 35% for services handling 10,000+ daily requests. Measurement note: These are historical results across my Digital Alpha services work.

Making generated output usable

I built in-chat matplotlib charts with S3 storage and 30-day Redis caching, then connected the output to the conversation UI. My Next.js/Redux work included multi-model history, streaming-state recovery, and provider-key management. The focus was the complete path from backend output to a user being able to retrieve, view, and continue working with it.

Shared data integrations

Alongside platform work, I implemented QuickBooks read connectors for 10 accounting resources, including invoices, bills, payments, and journal entries. These use Zod validation, normalized TypeScript models, pagination, modification-date filters, and retries through Nango. Stack: Django REST Framework, Next.js, Redux Toolkit, PostgreSQL, Redis, AWS S3, LangFlow, TypeScript, Nango. Explore the rest of my Digital Alpha work
All projects