We use analytics to understand how our website is used. No personal data is collected.

AM

Piyush Mishra

Senior Software Engineer — AI Platforms & Distributed Systems

10+ years shipping production backends, distributed systems, and AI platforms

I build production backend systems, distributed applications, and AI platforms across Python and TypeScript. Most recently a RAG research platform indexing 2M+ documents across 60 enterprise customers, and an OAuth integration framework covering 50+ SaaS platforms. IIM Indore alumnus with prior industrial operations experience at NTPC — engineering depth plus real domain expertise. Currently building RFQTrace and available for freelance work.

About Me

I’m a senior software engineer with 10+ years of professional experience shipping production backend systems, distributed applications, and AI platforms — and two years before that running operations and procurement at a 652MW power plant, which is why I build for manufacturers rather than about them.

I hold a Post Graduate Diploma in Management from IIM Indore and a B.E. in Mechanical Engineering from Madan Mohan Malviya Engineering College, Gorakhpur. I taught myself to code in 2016, with no CS degree and no bootcamp, and was tech lead on a 10-person team at that first engineering job by the time I left it.

I set technical direction and owned system design for products from 0 → 1 through scale, across backend, AI, and full-stack surfaces. I've led engineering teams of 5–20 engineers, mentored developers through code review and architecture guidance, and established CI/CD, testing, observability, and incident-readiness practices that stuck as team standards rather than one-off fixes.

Currently building RFQTrace solo — a multi-tenant RFQ and supplier quote management platform — and taking on a small number of freelance projects at a time.

The Path Here

Mechanical engineering and power plant procurement to procurement software — by way of an MBA, a marketing role, and teaching myself to code.

  1. 2006 – 2010

    B.E. Mechanical Engineering

    Madan Mohan Malviya Engineering College, Gorakhpur

    Mechanical engineering undergraduate — the discipline I later applied to plant operations, and the one I had to unlearn to write software.

  2. 2010 – 2012

    Operations Engineer

    NTPC Ltd

    Ran operations and procurement at a 652MW gas-fired power plant — ₹150 crore of vendor spend, daily fuel negotiations with GAIL and RLNG, and a 20-person team across four trades maintaining 22,000 pieces of plant equipment.

  3. 2012 – 2014

    Post Graduate Diploma in Management

    Indian Institute of Management, Indore

    Left engineering to study management full-time.

  4. 2014 – 2015

    Manager — Marketing

    VResolve Softech Pvt Ltd

    Product marketing and sales operations for a multi-platform product — budgets, forecasting, and joining sales calls to give technical depth.

  5. 2015 – 2016

    Taught myself to code

    Career transition

    Left marketing to learn JavaScript, React, and CSS full-time through self-directed study — no CS degree, no bootcamp.

  6. Jun 2016 – Oct 2020

    Senior Full-Stack Developer / Tech Lead

    Kintry, Delhi

    First engineering job. Designed end-to-end architecture, shipped 5 mobile apps reaching 50K+ downloads, built a WebRTC conferencing platform for 50 concurrent users, and led a cross-functional team of 10.

  7. Nov 2020 – Nov 2021

    Project Leader

    R Systems International, Noida

    Led a team of 5 across three client engagements — a white-label OTT video platform, a React Native invoicing app, and a Flutter/Python online training platform.

  8. Feb 2022 – Apr 2023

    Full-Stack Engineer

    Loot Discount Inc, Canada

    Serverless fintech infrastructure on AWS handling 500K+ API requests/month, with Stripe Connect and Plaid powering the full payment lifecycle.

  9. May 2023 – May 2025

    Sr. Full Stack Engineer

    Revfx, San Francisco

    AI-powered sales intelligence — a production RAG platform indexing 2M+ documents, a reusable OAuth library across 50+ SaaS platforms, and multi-step AI workflow orchestration.

  10. 2025 – 2026

    Independent

    Self-directed

    Published 17 long-form engineering articles, rebuilt mishhy.com, and researched the procurement tooling gap that became RFQTrace.

  11. Mar 2026 – Present

    Founding Engineer

    RFQTrace

    Back to procurement — this time building the software. Fifteen years after negotiating supplier quotes by hand, I'm building the tool I wanted then. Live at rfqtrace.com.

Skills & Expertise

All
Languages
Backend & APIs
AI / ML
Cloud & Infrastructure
Databases & Storage
Frontend & Mobile
Testing & Observability
Auth, Payments & Real-time

Python

Languages

90%

TypeScript

Languages

95%

JavaScript

Languages

95%

SQL

Languages

90%

Dart

Languages

80%

Java / Kotlin / Swift

Languages

75%

Node.js

Backend & APIs

95%

NestJS

Backend & APIs

90%

FastAPI

Backend & APIs

90%

Express.js

Backend & APIs

90%

REST APIs

Backend & APIs

95%

GraphQL

Backend & APIs

85%

Prisma ORM

Backend & APIs

90%

RAG Pipelines

AI / ML

95%

Vector Search (ChromaDB)

AI / ML

90%

LLM Integration (OpenAI, Anthropic, DeepSeek)

AI / ML

90%

Function Calling & Orchestration

AI / ML

88%

Prompt Engineering & Structured Outputs

AI / ML

90%

TensorFlow

AI / ML

80%

AWS (Lambda, API Gateway, SQS, SNS)

Cloud & Infrastructure

95%

Step Functions, Batch, Amplify, SES

Cloud & Infrastructure

90%

CloudFormation

Cloud & Infrastructure

90%

Docker

Cloud & Infrastructure

90%

CI/CD (CircleCI, Jenkins)

Cloud & Infrastructure

88%

Vercel & Heroku

Cloud & Infrastructure

85%

PostgreSQL

Databases & Storage

90%

MongoDB

Databases & Storage

88%

DynamoDB

Databases & Storage

88%

Firebase (Realtime DB & Firestore)

Databases & Storage

95%

Redis

Databases & Storage

88%

FaunaDB

Databases & Storage

85%

Redshift & Snowflake

Databases & Storage

85%

CosmosDB

Databases & Storage

85%

React

Frontend & Mobile

95%

Next.js

Frontend & Mobile

90%

Redux

Frontend & Mobile

88%

React Native

Frontend & Mobile

88%

Flutter

Frontend & Mobile

85%

Material UI

Frontend & Mobile

85%

Playwright

Testing & Observability

90%

Cypress

Testing & Observability

88%

Jest & Enzyme

Testing & Observability

88%

Detox (React Native E2E)

Testing & Observability

85%

Sentry

Testing & Observability

88%

Segment & FullStory

Testing & Observability

85%

Gleap

Testing & Observability

80%

Auth0

Auth, Payments & Real-time

95%

OAuth 2.0 (50+ platforms)

Auth, Payments & Real-time

95%

RBAC & Multi-tenancy

Auth, Payments & Real-time

90%

Stripe Connect

Auth, Payments & Real-time

90%

Plaid

Auth, Payments & Real-time

88%

WebRTC & Socket.io

Auth, Payments & Real-time

88%

Experience

  1. Founding Engineer

    RFQTrace

    Mar 2026 – Present

    Designed and built a multi-tenant RFQ and supplier quote management platform solo for small-to-mid-size manufacturers — the segment priced out of enterprise procurement suites. Live at rfqtrace.com.

    Key Achievements:

    • •Multi-tenant architecture with organisation-level data isolation and RBAC across 5 roles
    • •Offline-first caching (Dexie.js + Firebase Realtime DB) so the tool stays usable on low-connectivity plant floors and supplier visits
    • •Structured quote comparison on first-class fields — unit price, lead time, MOQ, payment terms — rather than free text
    • •One-link supplier portals that produce structured data instead of PDF attachments someone has to transcribe
    Next.js
    TypeScript
    Firebase Auth
    Firebase Admin SDK
    Dexie.js
    Vercel
  2. Independent Software Engineer

    Independent

    2025 – 2026

    Time between roles spent building independently and documenting a decade of production work in depth — the research that became RFQTrace.

    Key Achievements:

    • •Published 17 long-form engineering articles on RAG pipelines, OAuth patterns, multi-tenant architecture, and offline-first sync
    • •Designed and shipped mishhy.com — Astro with React islands, Firebase, full Playwright coverage and structured data
    • •Researched the procurement tooling gap that became RFQTrace
    Astro
    React
    TypeScript
    Firebase
    Playwright
  3. Sr. Full Stack Engineer

    Revfx

    May 2023 – May 2025

    Built AI-powered sales intelligence at a San Francisco SaaS company — a production RAG research platform, a reusable OAuth integration framework, and multi-step AI workflow orchestration. Owned frontend architecture, data infrastructure, and AI workflow delivery.

    Key Achievements:

    • •Built a RAG-based research copilot on ChromaDB and CosmosDB indexing 2M+ documents and processing 50K+ semantic searches/month across 60 enterprise customers; iterated on chunking strategy and retrieval latency for production throughput
    • •Reduced search infrastructure cost by 95% by replacing Algolia with ChromaDB-based vector search across 80 GB of data, while maintaining production SLA performance
    • •Cut a 300 GB Redshift warehouse by 40% using idempotent hashing to deduplicate ingested SaaS data, and reduced DB read load with a Redis caching layer
    • •Architected a reusable OAuth library covering 50+ SaaS integrations (Salesforce, Outreach, Salesloft, HubSpot, Segment, Gmail, Google Calendar), powering auth for 200 customer accounts and cutting new-integration build time by 60%
    • •Designed production AI workflows on OpenAI, Anthropic, and DeepSeek with structured outputs, function calling, and multi-step orchestration for sales lead generation, hiring signals, and lookalike account identification
    • •Delivered a production-ready frontend within 3 months and onboarded the first enterprise client; improved LCP by 40% through code splitting and asset optimisation, and established Cypress e2e testing, Sentry monitoring, and Segment/FullStory analytics
    NestJS
    FastAPI
    Prisma
    PostgreSQL
    RDS
    Redis
    Temporal
    Airflow
    Airbyte
    Redshift
    ChromaDB
    React
    Redux
    Auth0
    Cypress
    Sentry
    Segment
  4. Full Stack Engineer

    Loot Discount Inc

    Feb 2022 – Apr 2023

    Built serverless AWS infrastructure and the full payment lifecycle for a Canadian fintech SaaS — API Gateway, Lambda, SQS, SNS, Stripe Connect, and Plaid behind a React and Next.js frontend.

    Key Achievements:

    • •Designed a serverless architecture (API Gateway, Lambda, SQS, SNS) processing 500K+ API requests/month across 1,000+ users, and cut infrastructure cost 30% through CloudFormation automation
    • •Built FaunaDB APIs at sub-200ms query latency, orchestrated dev-to-prod data migration with AWS Step Functions, and managed a 50 GB analytical dataset in Snowflake
    • •Integrated Stripe Connect (registration, transfers, payouts, webhook events) and Plaid (bank and card connections, financial transaction webhooks)
    • •Improved catalogue search speed 3× (200%) with Algolia custom-tuned indexes, and improved frontend LCP by 40% through lazy loading and bundle optimisation
    AWS Lambda
    API Gateway
    SQS / SNS
    Step Functions
    CloudFormation
    Stripe Connect
    Plaid
    FaunaDB
    Snowflake
    Auth0
    Algolia
    React
    Next.js
    CircleCI
    AWS Amplify
  5. Project Leader

    R Systems International

    Nov 2020 – Nov 2021

    Led a team of 5 across three client engagements in Noida — a white-label OTT video distribution platform, a small-business invoicing app, and an online training platform.

    Key Achievements:

    • •dotstudioPro — built the AVOD/SVOD/TVOD white-label OTT frontend, increased streaming engagement by 25%, and cut new-partner turnaround 10% with e2e automation (Jest, Enzyme, Detox, CircleCI)
    • •Invoice Simple — refactored a production React Native codebase from class-based to hooks-based components, lifting the Play Store 5★ rating 10% and the App Store 8%, cutting the bug report rate 10% and improving maintainability 15%
    • •Learning People — delivered a web/Android/iOS app on a Flutter frontend and Python backend, reducing operational cost 15% by bringing launch forward 2.5 months with no additional headcount, and trained and mentored 2 team members
    React Native
    Flutter
    Python
    TypeScript
    Jest
    Enzyme
    Detox
    CircleCI
    Jenkins
  6. Senior Full-Stack Developer / Tech Lead

    Kintry

    Jun 2016 – Oct 2020

    Designed end-to-end architecture for Kintry's full platform in Delhi — 5 published Play Store apps, a WebRTC conferencing platform, sub-domain web stores, and ML inventory models. Led a cross-functional team of 10.

    Key Achievements:

    • •Shipped 5 mobile apps to Google Play reaching 50K+ downloads, and built a Node.js backend on Heroku and AWS Lambda — REST APIs, payments, live video, chat, and microservices
    • •Built a WebRTC-based conferencing platform supporting 50 concurrent users with sub-second signaling latency
    • •Built sub-domain web stores with WhatsApp and SMS ordering that achieved 10% weekly growth
    • •Implemented TensorFlow ML models for inventory management and operational decision-making
    • •Led and managed a cross-functional team of 10 across engineering, sales, and HR
    React
    Node.js
    Python
    Flutter
    Java / Kotlin
    Swift
    WebRTC
    Socket.io
    TensorFlow
    Firebase
    AWS Lambda
    Heroku
  7. Manager — Marketing → Career Transition

    VResolve Softech Pvt Ltd

    2014 – 2016

    Product marketing and sales operations for a multi-platform product (2014–2015), then transitioned to software engineering through self-directed study of frontend development — JavaScript, React, and CSS (2015–2016).

    Key Achievements:

    • •Owned marketing and sales operational planning — annual budget, expenditure scheduling, variance analysis, and corrective action
    • •Partnered with the product team to educate internal and external stakeholders on product capability
    • •Joined sales calls with prospects to provide technical depth on the product
    Product Marketing
    Sales Ops
    Budgeting
    Analytics
  8. Operations Engineer

    NTPC Ltd

    2010 – 2012

    Operations and procurement for a 652MW gas-fired power plant — the industrial procurement experience that RFQTrace is built on.

    Key Achievements:

    • •Managed vendor procurement worth ₹150 crore (~USD 25M) across plant operations
    • •Negotiated and managed daily fuel supply of 2–4.2 MMSCMD (~USD 590K/day) with GAIL and RLNG across multiple sourcing locations
    • •Led a 20-person team across electrical, mechanical, operations, and welding, maintaining 22,000 pieces of plant equipment
    • •Cut demineralised water wastage by 8% through benchmarking against other NTPC plants
    Industrial Procurement
    Vendor Negotiation
    Demand Forecasting
    Plant Operations

Portfolio Showcase

Explore interactive demos and visual previews of my recent projects, showcasing innovative solutions in AI, SaaS integrations, and full-stack development.

Featured Projects

RFQTrace

RFQTrace

Multi-tenant RFQ and supplier quote management SaaS. Procurement teams track RFQs, invite suppliers via one-link portal, compare quotes side by side, and award jobs — all without spreadsheets. Built from scratch with Next.js 16, Firebase RTDB, role-based permissions, and offline-first Dexie caching.

Next.js 16
TypeScript
Firebase RTDB
Dexie.js
Vercel
go-distributed-worker

go-distributed-worker

A distributed job execution platform built in Go — a smaller version of Celery + SQS + worker orchestration. Features REST API, priority queues, worker heartbeats with TTL-based liveness, job leasing with visibility timeouts, retry with exponential backoff, dead-letter queues, idempotency keys, scheduled and cron jobs, worker autoscaling, graceful shutdown, distributed locking, Prometheus metrics, and OpenTelemetry tracing.

Go
Redis
PostgreSQL
Docker
Distributed Systems
job-search-automation

job-search-automation

Aggregates remote roles ($50k+, senior-level) from Greenhouse, Lever, and Ashby APIs, drafts a tailored cover letter per job using a multi-provider LLM abstraction (Anthropic, OpenAI, Gemini, DeepSeek, OpenCode, Ollama), and puts everything in one spreadsheet for review. Does not auto-submit applications — the automation is in aggregation and drafting, not submission.

Python
LLM
Ollama
Streamlit
Docker
deepseek-codegen

deepseek-codegen

VS Code extension for DeepSeek code generation — integrates AI-powered code completion and generation directly into the editor workflow.

TypeScript
VS Code API
AI
OTT streaming platform

OTT streaming platform

High-performance video streaming platform frontend with categorized content, seamless playback, and white-label generation tooling for dotstudioPro partners.

React
Frontend
CSS/SCSS
Pelicargo

Pelicargo

Full-stack logistics management tool for cargo tracking and service coordination.

JavaScript
Fullstack
Node.js

Available for freelance projects

I take on a small number of projects at a time so I can focus properly on each. If you need a senior Next.js or Firebase developer who understands SaaS architecture, AI integration, and shipping production software — let's talk.

© 2025 Piyush Mishra. Built with Astro.js, TypeScript, and Framer Motion.