Ronit Mukherjee↓ resume
latent·space · portfolio

Ronit Mukherjee

AI Architect

Designs and ships production AI systems — multi-agent orchestration, RAG, LLMOps, GenAI. Backed by 13+ years of full-stack enterprise delivery.

stack · skills
LLMsGenAIRAGPrompt EngineeringNLPEmbeddingsLangChainAutoGenCrewAIMulti-agent orchestrationLLMOpsVector DatabasesAI Security / GuardrailsGCPAWSReactNext.jsReact NativeNode.jsGraphQLREST APIsMERN Stack
delivery flowstatic
requirements01
business problem
what needs AI
tool / data source
JIRA · GitHub · Slack · CRM
constraints
latency · cost · security
build02
orchestration
multi-agent · AutoGen · CrewAI
RAG
LangChain · embeddings · vector DB
guardrails
AI security · prompt hardening
LLMOps
evals · observability · deploys
ship03
SAW
Salesforce Agentic Workforce
SEAP
Enterprise Agentic Platform
Slack apps
Verizon · AWS integrations
commerce
B2C · Experience · CG Cloud

from business requirements to production-ready systems — inputs flow through designed components into shipped systems. marks the active stage.

13+
years shipping software
3
enterprise AI platforms shipped
20+
portfolio apps in progress
19
AI certifications
portfolio

apps shipped and in progress — work that earns trust before the resume does

RM GPT preview● live

RM GPT

Personal RAG chat assistant for documents

A Streamlit chat app that lets you upload PDF and Word documents, builds a FAISS vector index with HuggingFace embeddings, and answers questions using an Ollama LLM with retrieval-augmented generation.

stack
  • Streamlit
  • Python
  • LangChain
  • FAISS
  • Ollama
  • HuggingFace
AI TOOLS
● building

AI Task Runner

Autonomous agent that plans and executes dev tasks

A multi-agent CLI and web dashboard that breaks engineering tickets into steps, runs them with guardrails, and reports back. Built to make agentic workflows practical for real teams.

stack
  • Next.js
  • LangChain
  • OpenAI
  • Tailwind
AI TOOLS
● building

RAG Document Assistant

Chat with your documents using retrieval-augmented generation

Upload PDFs or paste URLs, chunk and embed them, then ask questions in natural language. Includes source citations and conversation memory.

stack
  • React
  • Python
  • LangChain
  • Pinecone
  • FastAPI
PRODUCTIVITY
● planned

Slack Ops Hub

Internal operations straight from Slack

A modular Slack app template for approvals, alerts, and on-call workflows. Designed for teams that live in Slack but need structured processes.

stack
  • Node.js
  • Slack API
  • Bolt
  • Redis
SAAS
● building

SaaS Starter Kit

Opinionated Next.js foundation for paid web apps

Auth, billing, teams, and admin scaffolding pre-wired so I can spin up new product ideas in hours instead of weeks.

stack
  • Next.js
  • TypeScript
  • Stripe
  • Prisma
  • PostgreSQL
AI TOOLS
● planned

Prompt Manager

Version, test, and deploy prompts like code

A lightweight prompt engineering workspace with A/B testing, version history, and one-click export to your LLM application.

stack
  • React
  • Node.js
  • MongoDB
  • OpenAI
OPEN SOURCE
● planned

Dev Tools Directory

Curated, filterable directory of tools developers actually use

A community-driven directory with category filters, search, and stack tags. Built to stay fast even with hundreds of listings.

stack
  • Next.js
  • Tailwind
  • MDX
overview

who I am, in three paragraphs

AI Architect and full-stack solution leader with 13+ years designing and delivering scalable enterprise applications and AI-ready platforms. Hands-on across GenAI/LLMs, RAG, prompt engineering, LLMOps, embeddings, vector databases, Python, AWS, and GCP — with deep delivery credibility in React, Node.js, and the MERN stack.

I translate business needs into production AI systems: multi-agent orchestration platforms, RAG pipelines, guardrailed LLM apps, and the LLMOps around them. I lead cross-functional teams, mentor engineers, and own outcomes end-to-end — from solution architecture through deployment and measurable impact.

Currently focused on agentic systems, LLMOps, and shipping products that combine AI with practical engineering.

capabilities

engineering capabilities — what I work with

GenAI / LLMs

  • LLMs
  • GenAI
  • RAG
  • Prompt Engineering
  • NLP
  • Embeddings

Agent frameworks

  • LangChain
  • AutoGen
  • CrewAI
  • Multi-agent orchestration

LLMOps & infra

  • LLMOps
  • Vector Databases
  • AI Security / Guardrails
  • GCP
  • AWS

Full stack

  • React
  • Next.js
  • React Native
  • Node.js
  • GraphQL
  • REST APIs
  • MERN Stack

Practices

  • Solution architecture
  • Cross-functional leadership
  • Tech mentorship
deployments

13+ years, reverse-chronological — what I shipped, where

  1. Jun 2022
    Present

    Technical Architect — AI, Slack & MERN

    current
    Salesforce
    • Architected and delivered SAW (Salesforce Agentic Workforce), a multi-agent orchestration platform that turns tools like Cursor and Claude into a reusable enterprise agent workforce — accelerating AI workflow delivery by 40%.
    • Designed and launched SEAP (Salesforce Enterprise Agentic Platform), integrating JIRA, GitHub, and reusable agent pipelines to standardize development workflows and reduce manual coordination by 30%.
    • Led custom Slack app development and Slack API integrations for enterprise customers including Verizon and AWS, automating recurring workflows across two large accounts.
    • Led frontend delivery for Salesforce B2C Commerce Cloud, Experience Cloud, and Consumer Goods Cloud; built React Native, React, and Next.js applications for internal and customer-facing use cases.
  2. Sep 2018
    May 2022

    Senior Associate, Experience Technology L2

    Publicis Sapient · Noida
    • Led development of scalable enterprise applications, improving system performance by 30% and supporting more responsive user experiences.
    • Implemented new frameworks and delivery practices, reducing deployment time by 40% across key project releases.
    • Worked with client stakeholders to refine requirements and align technical solutions with business goals, improving delivery accuracy and project outcomes.
  3. Dec 2015
    Sep 2018

    Senior Software Engineer

    Icreon Communications Pvt Ltd · Noida
    • Developed scalable web applications using React and Node.js, improving user engagement by 30% and strengthening frontend performance.
    • Optimized backend services and refactored code paths, reducing server load time by 20% while improving overall application reliability.
    • Collaborated with cross-functional teams and clients to deliver projects on time and within scope, ensuring high code quality and smooth deployment.
  4. Oct 2015
    Dec 2015

    Senior Software Engineer

    MyOperator Voicetree Technologies · New Delhi
    • Designed and delivered scalable VoIP features that improved communication efficiency and supported faster issue resolution for production users.
    • Worked with cross-functional teams to integrate new capabilities and stabilize service behavior, improving product reliability during active releases.
    • Resolved complex technical issues quickly, helping maintain service quality and minimizing downtime for customer-facing systems.
  5. Jan 2013
    Sep 2015

    Software Engineer

    Medma Infomatix Pvt Ltd
    • Developed and maintained web applications, improving load times by 30% and enhancing overall user experience.
    • Built scalable solutions with modern coding practices, increasing system stability and reducing production issues.
    • Collaborated with team members to gather requirements and deliver projects on time, supporting consistent release quality.
  6. May 2010
    Apr 2012

    Trainer

    Computer Informatix Center
    • Delivered PHP and web development training to graduate learners, covering core concepts through advanced application building for 20+ trainees per batch.
    • Created a structured curriculum and hands-on workshops that improved coding and problem-solving skills across the program.
    • Tracked learner progress and provided feedback to strengthen outcomes, helping trainees build production-ready technical foundations.
case studies

three featured systems — problem, stack, outcome

01

SAW — Salesforce Agentic Workforce

Multi-agent orchestration platform

problem

Enterprise teams were paying per-seat for individual AI tools (Cursor, Claude) without a way to reuse agent work across projects or enforce shared guardrails.

outcome

Architected SAW as a reusable agent workforce — turning one-off AI tool usage into standardized, governed agent pipelines that enterprise teams could compose across workflows.

stack
  • Multi-agent orchestration
  • AutoGen
  • CrewAI
  • Cursor
  • Claude
  • Slack API
02

SEAP — Salesforce Enterprise Agentic Platform

Agent pipelines wired into dev workflows

problem

Development coordination across JIRA, GitHub, and agent pipelines was manual — context was lost between steps and agent runs couldn't be standardized.

outcome

Designed and launched SEAP, integrating JIRA and GitHub with reusable agent pipelines to standardize development workflows end-to-end.

stack
  • JIRA integration
  • GitHub integration
  • Agent pipelines
  • LLMOps
03

Slack apps for Verizon & AWS

Enterprise Slack integrations at scale

problem

Two large enterprise accounts needed recurring workflows automated inside Slack without breaking their existing security and compliance posture.

outcome

Led custom Slack app development and Slack API integrations for Verizon and AWS, automating recurring workflows and improving collaboration across both accounts.

stack
  • Slack API
  • Node.js
  • Salesforce
  • Enterprise auth
benchmarks

measurable outcomes — verifiable from the work above

40%
faster AI workflow delivery
SAW · Salesforce
30%
less manual coordination
SEAP · Salesforce
30%
improved system performance
Publicis Sapient
40%
faster deployments
Publicis Sapient
30%
better user engagement
Icreon
20%
reduced server load
Icreon
30%
faster page loads
Medma Infomatix

Every number above is drawn directly from the work where it was delivered. Source projects and positions are listed in deployments and case studies.