AI MANAGER · ML ENGINEER · APPLIED AI

Shubhabrata
Ganguly

I build intelligent, production-grade AI systems — from ML & data science to LLM, RAG & agentic AI — and lead the product and quality work that ships them. 11+ years across enterprise engineering at PwC, Deloitte & Infosys.

📍 Kolkata, India · 🌐 Remote & Hybrid · Open to roles & freelance
11+ yrs experience
15engineers led
4service pillars
8+ certifications
Portrait of Shubhabrata Ganguly
LLM · RAG · MCP
Agentic AI
MLOps
01 — About

Engineer at the core, AI by design.

I'm a results-driven engineering leader with 11+ years building scalable automation frameworks, cloud-integrated pipelines and API-driven systems across enterprise environments. I'm now channeling that foundation into AI/ML engineering — currently completing a Post Graduate Program in AI & Machine Learning from The University of Texas at Austin (McCombs / Great Learning).

My focus sits at the intersection of LLM engineering, Generative AI, RAG architectures, agentic systems and the Python AI ecosystem — combined with the product sense and quality discipline needed to turn prototypes into dependable products.

I'm available for full-time roles and take on freelance engagements across the four pillars below. Whether you need a model built, an AI application shipped, a roadmap defined, or a quality strategy for an AI product — let's talk.

Microsoft Azure AI Fundamentals AWS Cloud Practitioner SAFe® 6 POPM CSPO® SAFe 6 Agilist Lean Six Sigma UiPath Test Automation JMeter Performance
02 — Services

Freelance & consulting, across four pillars.

End-to-end delivery or focused engagements — pick a pillar, or combine them for idea-to-production AI.

01 📊

ML & Data Science

From raw data to models that drive decisions — statistically sound and business-aligned.

  • Machine Learning & model development
  • Data science & exploratory analysis
  • Financial data analysis
  • Data modelling & feature engineering
  • KPI analysis & metric design
  • Competition & market benchmarking
02 🤖

Applied AI

Generative AI and agents built to run reliably in production, not just in demos.

  • LLM engineering & prompt design
  • RAG pipelines & vector search
  • MCP & tool-augmented AI
  • Building with Claude & other frontier models
  • Agentic AI & AI agent building
  • AI application building
  • MLOps · AIOps · model deployment
03 🧭

Product Management

Turning AI capability into products with a clear roadmap and measurable outcomes.

  • AI Product Management (AI PM)
  • Roadmap design & prioritization
  • Product strategy & discovery
  • KPI & success-metric definition
  • Competitive & market analysis
  • Agile / SAFe & stakeholder management
04

Quality Analysis

15 years of quality DNA — now applied to both classic software and AI systems.

  • AI QA & model evaluation
  • Automation testing (Selenium, Playwright, Appium)
  • Software / functional testing
  • Performance & load testing (JMeter)
  • CI/CD quality gates
  • Test strategy & frameworks

Not sure which pillar you need? Describe your problem and I'll scope it.

Request a proposal
03 — Experience

Where I've delivered.

Software Engineering Manager — Quality & Automation

Jul 2024 — Present

PwC Acceleration Centre

  • Directed a 15-member engineering team delivering enterprise automation across web, mobile & API layers.
  • Championed AI-assisted tooling for code review and test generation — 35% faster onboarding.
  • Embedded quality gates into Azure CI/CD; introduced Playwright, lifting execution speed ~40%.

Senior Automation Engineer

Jan 2021 — Jul 2024

PwC Acceleration Centre

  • Architected a reusable Java + Selenium + REST Assured framework — 50% shorter regression cycles.
  • Built Azure Pipelines CI/CD for nightly cross-platform regression with zero manual effort.
  • Led iOS/Android mobile automation with Appium; mentored engineers on framework design.

Software Engineer — Automation & Performance

Aug 2018 — Oct 2020

Deloitte USI

  • Engineered end-to-end mobile automation with Appium across Android & iOS.
  • Led performance workstreams (JMeter, BlazeMeter), modelling real-world load to find bottlenecks.

Software Engineer

Feb 2014 — Jun 2018

Infosys Ltd.

  • Built test plans, automation scripts & reusable utilities for large-scale enterprise projects.
  • Drove root-cause analysis and testability reviews influencing architecture decisions.
04 — Projects

Selected work & builds.

A mix of open-source builds and applied AI/ML projects. More on GitHub ↗.

Agentic AITypeScript

Multi-Agent Business

Spin up a business with 38 collaborating AI agents — orchestration, roles and workflows.

View repo ↗
ML & DSMLOps

SuperKart Sales Forecasting

Sales prediction service — Flask API backend + Streamlit frontend, fully Dockerized.

View repo ↗
Applied AIRAG

Open Notebook AI

Open-source take on NotebookLM — flexible document intelligence and retrieval features.

View repo ↗
GenAITypeScript

ImageGen

A generative image application exploring text-to-image workflows.

View repo ↗
Applied AIPython

AudioBook & Translator

Converts text to audiobooks with translation — speech + language processing pipeline.

View repo ↗
ML & DSHands-on

AIML PGP — Hands-on

Coursework & hands-on notebooks from the UT Austin AI/ML program.

View repo ↗
Applied projects from practice: AI-Powered Test Automation Assistant (LangChain + OpenAI → auto-generates Selenium scripts), RAG Knowledge Retrieval (LlamaIndex + FAISS + GPT-4), and API Performance Anomaly Detection (Isolation Forest on JMeter logs).
05 — Skills

The toolkit.

AI / ML & GenAI

PythonNumPyPandasScikit-learn TensorFlowPyTorchDeep LearningNLP LLM Fine-tuningRAGLangChainLlamaIndex OpenAI APIClaudeHugging FaceFAISS ChromaDBEmbeddingsAI AgentsMCP

Engineering & Cloud

JavaJavaScriptBashDocker AzureAWSAzure DevOpsGitHub Actions JenkinsCI/CDREST APIsMLOpsAIOps

Quality & Automation

SeleniumPlaywrightAppiumREST Assured JMeterBlazeMeterPerformance Testing API AutomationAI QA

Product & Delivery

Product StrategyRoadmappingKPI Analysis Market ResearchAgileSAFe 6.0Scrum Stakeholder MgmtLean Six Sigma
06 — Contact

Let's build something intelligent.

Hiring, or have a project in mind? I reply within a day. Tell me about your role or the problem you're solving and I'll come back with a clear next step.