Enterprise AI systems, product-grade architecture, and conversational platforms

Mukesh Lohumi

AI Engineer & Solution Architect

AI Engineer and Solution Architect focused on LLM systems, agentic automation, conversational AI, and cloud-ready backend engineering.

View Projects Resume

Agentic

AI Systems

Retrieval

RAG Pipelines

Delivery

Cloud-Native

~7 YearsAI and Software Engineering Experience

8+ ProjectsEnterprise AI, RAG, Chatbots & Backend Systems

NIT SikkimB.Tech in Computer Science Engineering

About

I operate at the intersection of AI engineering, backend systems, enterprise solution design, and product building.

Mukesh Lohumi

I build production-minded AI systems that connect language models, retrieval, APIs, cloud infrastructure, and user-facing product experiences. That range lets me move from architecture decisions to backend implementation to the final workflow that a team or customer actually uses.

Education

National Institute of Technology Sikkim

Bachelor of Technology in Computer Science Engineering

August 2015 - May 2019

AI Systems

Production-oriented LLM applications, agentic workflows, tool calling, prompt design, and orchestration for real business processes.

RAG & Knowledge Retrieval

End-to-end retrieval pipelines covering ingestion, chunking, embeddings, indexing, filtering, and high-accuracy assistant behavior.

Chatbots & Automation

Enterprise virtual assistants, CRM-connected conversational workflows, WhatsApp automation, and multi-step customer journeys.

Backend & Cloud Architecture

API design, microservices, caching, observability, CI/CD, and cloud-native delivery using Docker, Kubernetes, Azure, and AWS.

What I'm building and exploring

LLM-powered enterprise assistantsAI workflow automation systemsWhatsApp-based marketing and engagement productsJob-tech and platform productsAI-powered screen and UI generation toolsScalable retrieval and knowledge systems

Core Expertise

Built around enterprise AI delivery, architecture thinking, and product execution.

The skill map is structured to show how I operate: from LLM systems and conversational workflows to backend delivery, cloud infrastructure, and enterprise integrations.

Core expertise

How I create leverage across product and engineering

AI / LLM

LLM Application DevelopmentAgentic AI SystemsRetrieval-Augmented Generation (RAG)Prompt EngineeringTool CallingWorkflow OrchestrationMulti-Agent Orchestration

Conversational Systems

Enterprise Chatbot DevelopmentConversational AINLP Workflow DesignWhatsApp AutomationCustomer Support AutomationLead Qualification Workflows

Architecture / Product

Solution ArchitectureBackend ArchitectureAPI IntegrationsCRM IntegrationsSaaS / Product PrototypingEnterprise Workflow Design

Backend

Technical stack

Node.jsExpressPythonFastAPIDjangoFlaskNext.jsReact.js

AI / Conversational Stack

Technical stack

GPT-4GPT-4oAzure OpenAILangChainLangGraphLangSmithAzure AI SearchAzure AI FoundryMicrosoft CopilotCopilot StudioDialogflowOpenAI

Cloud / DevOps

Technical stack

AWSAzureDockerKubernetesAzure DevOpsAzure FunctionsCI/CDRedisElasticsearch

Data / Integrations

Technical stack

PostgreSQLMongoDBMySQLSAP API IntegrationsWhatsApp Business APIsHubspotShopifyWooCommerceRazorpayPaytm

Featured Projects

Case-study style work across enterprise AI, retrieval systems, automation, analytics, and backend platforms.

Each project is framed around the problem, the system that was built, and the capability it unlocked.

Agentic AI

Enterprise Agentic AI Solutions

Enterprises needed AI systems that could do more than answer questions. They needed workflows that could reason across steps, execute tools, and fit into existing business operations.

What I built

Built enterprise-grade agentic systems using LangChain, LangGraph, Azure OpenAI, and GPT-4/4o with tool calling, memory, validations, orchestration, and autonomous task execution.

LangChainLangGraphAzure OpenAIGPT-4/4oFastAPINode.js

Outcome

Delivered production-oriented AI systems designed for enterprise workflows, not isolated prototypes.

RAG Systems

Enterprise RAG Virtual Assistant

Knowledge-heavy enterprise environments need assistants that retrieve the right information reliably, not just generate plausible responses.

What I built

Designed scalable retrieval pipelines spanning ingestion, chunking, embeddings, Azure AI Search indexing, metadata filtering, and LLM-powered retrieval for virtual assistants and knowledge systems.

Azure AI SearchAzure OpenAIRAGEmbeddingsPythonFastAPI

Outcome

Created retrieval-driven assistants focused on answer quality, context control, and enterprise knowledge access.

Automation

CRM and WhatsApp Automation Systems

Sales, support, and operations teams often rely on fragmented tools that create slow follow-up loops and manual process gaps.

What I built

Built conversational and workflow automation systems connected to CRMs, calendars, and WhatsApp Business APIs for lead qualification, support automation, and process execution.

WhatsApp Business APIsCRM IntegrationsCalendarsNode.jsAzure Functions

Outcome

Turned messaging and workflow touchpoints into integrated automation systems with clear operational value.

Conversational Commerce

Conversational Commerce / Chatbot Integrations

Customer-facing chatbot experiences need strong conversation design and reliable business integrations to move beyond scripted flows.

What I built

Built chatbot journeys and reusable components on Yellow.ai, integrating payment gateways and commerce systems including Razorpay, Paytm, Hubspot, Shopify, and WooCommerce.

Yellow.aiNode.jsRazorpayPaytmHubspotShopifyWooCommerce

Outcome

Delivered end-to-end conversational experiences that connected product logic, operations, and client delivery.

Analytics Products

Social Media Analytics Dashboard

Teams needed a cleaner way to understand cross-platform performance without stitching together disconnected channel reports.

What I built

Developed analytics dashboards using YouTube, Facebook, and Twitter APIs with Node.js and React.js to surface business-facing metrics and platform insights.

Node.jsReact.jsYouTube APIFacebook APITwitter API

Outcome

Shipped data products that combined API engineering, backend aggregation, and clear insight presentation.

Cloud-Native Backend

Society Management Backend Platform

Operational platforms need dependable backend services that are deployable, maintainable, and ready for production scale.

What I built

Developed backend services for a society management platform using microservices architecture, Docker, Kubernetes, AWS, and CI/CD pipelines.

Node.jsMicroservicesDockerKubernetesAWSCI/CD

Outcome

Strengthened backend engineering foundations across service design, orchestration, and production deployment.

Applied ML

Ad Fraud Detection System

Digital ad ecosystems are vulnerable to invalid activity, and detecting fraud requires both practical heuristics and machine learning insight.

What I built

Built a rule-based and machine-learning-driven ad fraud detection system using Python, Flask, and MySQL, contributing to recognized research work.

PythonFlaskMySQLMachine LearningPandas

Outcome

Combined product execution with applied ML research, leading to IEEE conference recognition.

Virtual Assistants

Intelligent Virtual Assistant

Practical virtual assistants need more than intent matching. They must translate conversation design into useful product experiences.

What I built

Developed an intelligent virtual assistant using Google Dialogflow, Node.js, and React, combining conversational UX with practical product development.

DialogflowNode.jsReact.jsConversational UX

Outcome

Delivered a working assistant experience that connected conversational logic with product execution.

Experience

A journey across AI systems, analytics platforms, conversational products, and startup execution.

The experience timeline is written to show technical depth and business-facing delivery together: architecture, implementation, integrations, and production outcomes.

May 2023 - Present

AI Engineer

Dsmatics (AccelEdge AI)Remote

Building production AI systems for enterprise use cases across retrieval, automation, CRM workflows, and conversational experiences.

Delivered enterprise-grade agentic AI solutions using LangChain, LangGraph, Azure OpenAI, and GPT-4/4o

Built scalable RAG pipelines with ingestion, chunking, embeddings, Azure AI Search indexing, and metadata filtering

Developed multi-agent workflow automation and Copilot-integrated enterprise assistants

Shipped LLM backends with Python, FastAPI, Node.js, Azure Functions, Redis, logging, and observability

Containerized and deployed systems with Docker, Kubernetes, and Azure DevOps CI/CD

May 2022 - May 2023

Software Developer

Dsmatics (AccelEdge AI)Remote

Built data-driven dashboards and cloud-native backend systems, strengthening production engineering across analytics, microservices, and deployment.

Developed analytics dashboards using YouTube, Facebook, and Twitter APIs with Node.js and React.js

Built backend services for a society management platform using microservices architecture

Containerized services with Docker and orchestrated deployments through Kubernetes

Delivered cloud deployments on AWS using CI/CD pipelines

Aug 2020 - May 2022

Software Engineer

Yellow.aiBengaluru, India

Delivered end-to-end conversational solutions by combining chatbot logic, reusable development patterns, client communication, and production deployment.

Created conversational chatbots and designed user journeys with NLP logic

Built reusable components using Node.js on the Yellow.ai platform

Integrated CRMs and payment systems including Razorpay, Paytm, Hubspot, Shopify, and WooCommerce

Handled full delivery from client interaction and logic design to production launch

Jan 2018 - Aug 2020

Co-Founder

Dezinee.aiUttarakhand, India

Combined engineering, product building, and startup execution to deliver practical ML and conversational products while leading operations and growth strategy.

Built a rule-based and ML-driven ad fraud detection system using Python, Flask, and MySQL

Developed an Intelligent Virtual Assistant using Dialogflow, Node.js, and React

Led product development, team operations, client communication, and business growth strategy

IEEE SITIS 2019

Research paper accepted — Digital Advertising Fraud Detection System, Italy

Contact / Opportunities

Open to AI engineering, backend architecture, chatbot platforms, and product-building opportunities.

If you are building enterprise AI, conversational products, workflow automation, or backend-heavy systems, I can help bridge architecture, implementation, and product delivery.

Available for recruiter conversations, product teams, startup collaborations, and enterprise solution discussions.

Email Me LinkedIn GitHub Download Resume

Mukesh Lohumi

AI Engineer

Available for opportunities

Primary contact

lohumi.mukesh@gmail.com

Best fit

Enterprise AI platforms, LLM products, workflow automation, chatbot ecosystems, backend-heavy SaaS products, and architecture-led engineering roles.