~/rameshsnotebook — zsh — 120×40
ramesh@rameshsnotebook ~ $ whoami --verbose

15+ years engineering
5 industries shipped
Cloud-native systems
Full stack development
Production-grade Multi-Agentic AI
01 About // who.md

15 years shipping software — Java, Spring Boot, microservices, cloud. The last few deep in LangGraph, CrewAI, RAG pipelines and multi-agent systems.

profile.json
const ramesh = {
  name: "Ramesh Maharaddi",
  site: "rameshsnotebook.com",
  role: "Senior Software Engineer & Solution Designer",
  focus: ["Agentic AI", "LangGraph", "CrewAI", "RAG", "LLMs"],
  experience: 15,
  domains: ["Telecom", "Banking", "Insurance", "Retail", "Fintech"],
  passion: "making AI reliable, observable, and scalable",
  currentlyBuilding: "multi-agent platforms with MCP integration"
};
02 Skills & Tools // 120+ technologies

The stack I reach for when building production systems, not demos.

Agentic AI 08
LangChainLangGraphCrewAIAutoGenSpring AISemantic KernelLlamaIndexHaystack
LLM Platforms 08
Azure OpenAIAWS BedrockVertex AIOllamavLLMPortkeyLiteLLMMCP
RAG & Vector DB 08
PineconeQdrantWeaviatePGVectorFAISSChromaAzure AI SearchGraphRAG
Doc Intelligence 07
PyMuPDFUnstructured.ioAzure DIGPT-4o VisionTesseractLayout ParserApache Tika
Backend 08
JavaPythonTypeScriptSpring BootSpring CloudFastAPIExpress.jsgRPC
Frontend 07
ReactAngularNext.jsTailwindAnt DesignReduxWebSockets
Data & Pipelines 08
Apache KafkaPySparkAirflowRayApache CamelDatabricksDelta LakeMS Fabric
Cloud & DevOps 08
AWSAzureGCPKubernetesDockerHelmArgo CDTerraform
03 AI Tech Stack // production-grade

How I layer an AI system from reasoning through to deployment.

Reasoning Layer

  • GPT / Claude / Gemini
  • Llama / Mistral / DeepSeek
  • Function calling
  • Structured outputs
  • Reasoning chains

Orchestration

  • LangGraph state machines
  • CrewAI role-based crews
  • MCP tool integration
  • Supervisor patterns
  • Human-in-the-loop

Knowledge Layer

  • Hybrid BM25 + vector
  • Re-ranking pipelines
  • GraphRAG
  • Embedding fine-tuning
  • Source attribution

Observability

  • LangSmith tracing
  • Prompt versioning
  • Evaluation harnesses
  • Cost & latency metrics
  • Drift detection

Safety & Ops

  • Guardrails AI
  • Prompt injection defense
  • PII redaction
  • Rate limiting / quotas
  • Fallback strategies

Infrastructure

  • Kubernetes + Helm
  • Argo CD GitOps
  • Azure / AWS / GCP
  • Vector DB clustering
  • Multi-region failover
04 My Projects // selected

Systems I've designed and shipped end to end, in production environments.

~/projects/agentic-platform FEATURED
Agentic AI Orchestration Platform
Multi-agent system built with LangGraph and CrewAI where agents handle research, analysis, execution and review tasks. A supervisor agent coordinates everything, with tools wired in via MCP.
LangGraphCrewAISpring AIMCPAzure OpenAIPinecone
~/projects/rag-pipeline PROD
RAG Knowledge Pipeline
Processes PDFs, PPTX and Excel files using hybrid BM25 + vector search with re-ranking. Built for large-scale document retrieval in enterprise environments.
LlamaIndexQdrantPyMuPDFAzure DIKafka
~/projects/knowledge-assistant LIVE
Enterprise Knowledge Assistant
Conversational assistant backed by RAG with source attribution and hybrid search. LangSmith tracing in production. Running for thousands of users daily.
LangChainPGVectorFastAPIReactLangSmith
~/projects/doc-intelligence PROD
Document Intelligence Engine
Parses scanned PDFs, invoices and contracts with layout-aware extraction. Uses GPT-4o Vision and Azure Document Intelligence to handle messy real-world documents.
PyMuPDFAzure DIGPT-4oLlamaIndexPython
~/projects/reward-platform SHIPPED
Enterprise Reward Platform
Corporate order management platform built to TM Forum Open API standards. Spring Boot microservices on Kubernetes with Kafka handling the event flow.
Spring BootKafkaK8sReactTM Forum
~/projects/gov-travel SHIPPED
Government Travel Portal
Travel verification system for a national government handling cross-border entry. Angular frontend, Spring Boot backend, with strict security and compliance requirements throughout.
AngularSpring BootAWS AuroraOAuth2
05 Public Websites // built & published for community

Side projects I built and put out there for free. Useful tools, nothing behind a paywall.

mydevtools.in

87+ developer utilities, all running in the browser. Format, diff, decode, generate — your data never leaves your machine.

tinycut.in

URL shortener with dynamic links, QR codes, bio pages and click analytics. Tracks location and referral source, with team access support.

belgaum.today

News site covering Belgaum and nearby areas. Timely local coverage without the noise.

appliedagentic.in

Free learning resource on agentic AI, LLMs and multi-agent systems. Covers the basics through to how things actually work in production.

06 AI Passions // what excites me

The problems in AI I find genuinely interesting to work on.

⚡
Agentic AI Systems

Building agents that reason through a problem, take action and self-correct. The challenge is making them reliable in real systems, not just demos.

🕸️
Multi-Agent Orchestration

Getting multiple agents to work together without stepping on each other is genuinely hard. I enjoy solving the coordination problems.

🔍
RAG Systems

Getting RAG to work well in practice is harder than it looks. Chunking strategy, hybrid retrieval, re-ranking — these details matter a lot.

🔌
Model Context Protocol

MCP is the right way to connect agents to tools. I build and integrate MCP servers so agents get structured, safe access to real systems.

📡
Agent-to-Agent Protocol

How agents should talk to each other across systems is still being figured out. I work on communication patterns that hold up under real load.

🧬
Knowledge Engineering

GraphRAG, tuned embeddings and structured knowledge graphs that make complex search actually work over messy enterprise data.

🤖
Agentic Chatbot

Chat assistants that actually do things — plan steps, call tools, remember context. Not just a thin wrapper around an LLM API.

🎤
Voice Bot

Real-time voice interfaces where speech, LLM reasoning and text-to-speech all have to work together with low latency. Latency is everything here.

🔒
AI Safety & Reliability

AI that fails gracefully beats AI that fails hard. Guardrails, injection defense, fallback paths and eval harnesses matter once you're in production.

📋
AI Governance

Making AI auditable and explainable is part of the job. Audit trails, policy controls and responsible practices aren't optional in serious deployments.

07 Experience // 15+ years

Roles I've held and what I actually built in each one.

JAN 2022 — PRESENT · SINGAPORE · CURRENT
Singapore Telecommunication (Singtel)
Senior Microservice Developer & Solution Designer
Leading AI platform design at Singtel — building multi-agent workflows, shared RAG services and LLM gateway integrations across enterprise systems.
  • Design and build multi-agent systems using A2A, ACP and MCP patterns for enterprise automation.
  • Build and maintain RAG pipelines in production using hybrid retrieval, chunking strategies and re-ranking across vector stores.
  • Led solution design for the corporate order management platform using TM Forum Open APIs. Own end-to-end delivery from requirement through to production.
  • Build LLM-integrated apps and document ingestion pipelines covering PDFs, Excel, PPTX and scanned files.
  • Build event-driven microservices and deploy them on Kubernetes and OpenShift using GitOps workflows.
  • Handle security: OAuth2/JWT, secrets management, PII redaction and guardrail evaluation.
  • Mentor junior engineers and run architecture reviews, design sprints and production postmortems.
NOV 2019 — DEC 2021 · SINGAPORE
DXC Technology Services Singapore
Svc Info Developer III
Led development of the SafeTravel Portal for the Singapore Government. Full-stack app for cross-border travel verification with strict security and compliance requirements.
  • Built the frontend, microservices backend and database layer. Set up ETL pipelines for batch document processing.
  • Owned code quality, CI/CD, team workload and production deadlines.
  • Set up SSO and OAuth2 access control to meet government compliance standards.
JUL 2016 — AUG 2019 · BANGALORE
Wells Fargo India Solution
Senior Technology Specialist
Built backend components for global banking workflows at Wells Fargo.
  • Built REST APIs, event-driven integrations and data pipelines. Worked with stakeholders across multiple time zones.
  • Drove API standardisation and led the microservices migration across banking systems.
SEP 2014 — MAR 2016 · BANGALORE
J.P. Morgan Services India
Application Developer
Built financial applications and batch integration components for J.P. Morgan's trading and settlement systems. Performance and reliability were non-negotiable.
NOV 2010 — AUG 2014 · BANGALORE
TechTree IT & Accenture
Software Engineer Analyst
Delivered web applications across Insurance and Retail at TechTree and Accenture. Mentored junior developers and built a reputation for clean code and shipping on time.
08 Certifications // continuous learning

Formal learning that runs alongside 15 years of hands-on production work.

🏆
Agentic AI Professional Certification
NVIDIA
Expert
🎓
Applied Agentic AI for Organizational Transformation
MIT Professional Education
Strategic
☸️
KCNA — Kubernetes & Cloud Native Associate
Linux Foundation / CNCF
Associate
🤖
Multi-AI Agent Systems with CrewAI
DeepLearning.AI
Specialist
☕
Oracle Certified Associate — Java Programmer
Oracle
Professional
🔌
API Design & Security on Apigee (×3)
Google Cloud / Coursera
3 Certs
09 Connect // let's build

Open to interesting projects, consulting work, and senior AI roles.

Let's work together

Looking for interesting projects, consulting work, or senior roles at companies doing real things with AI. Remote or hybrid preferred.

→ LinkedIn → GitHub
also on

Send a Message