The USB-C for AI: a complete guide to the Model Context Protocol
Understand MCP through a restaurant analogy, then build, connect and deploy your own server from scratch.
FIG. 01 · THE STACK
KARTHIKDILEEP
- experience2 yrssince Jun 2024
- companies2Devmani · S2 Data
- systems shipped4infra · fintech · AI
- current focusAgentic AILangChain · Agno · MCP
background · education · how I work
01
About
I'm a Full Stack Engineer with two years of experience building scalable applications using Python, FastAPI, JavaScript, React, PostgreSQL, AWS and GCP.
My work spans microservices, distributed systems and API development. Lately that has meant designing agentic-AI applications with LangChain, Agno, MCP servers and multi-agent orchestration workflows.
I care about systems that stay reliable, performant and observable under load: asynchronous backends, automated data pipelines, infrastructure-as-code, and interfaces that are a pleasure to use.
I hold a B.Tech in Computer Science Engineering from MIT ADT University, Pune (2020–2024).
grouped by layer · languages to infrastructure
02
Skills & Stack
01 Languages
03 · core
02 Backend
06 · services
03 Frontend
03 · interfaces
04 Data & Storage
04 · persistence
05 Cloud & Infra
06 · platform
06 AI & Agents
07 · intelligence
07 Automation
03 · tooling
08 Practices
05 · methods
jun 2024 - present · two companies
03
Experience
Full Stack Engineer
Devmani Traders Pvt. Ltd.
Wealthh, a mutual fund investment platform, and Kalman, its AI investment assistant
- Developed Kalman, an AI investment assistant on an Agno multi-agent system, routing each query to specialist agents for fund research, portfolio reviews and orders.
- Built MCP servers exposing fund data, portfolio analytics and BSE orders as agent tools, with a hybrid LLM and embedding-based tool selector.
- Architected the FastAPI and PostgreSQL backend with the BSE StAR MF order integration, shipped via GitHub Actions CI/CD with test gates and blue/green deploys on GCP.
Associate Software Engineer
S2 Data Systems Pvt. Ltd.
SyncSafe, an AWS disaster-recovery and infrastructure-as-code platform
- Engineered a Pulumi provisioning engine deploying AWS workloads with automatic rollback, plus cross-region disaster-recovery sync in boto3 with drift detection.
- Developed the FastAPI backend on PostgreSQL and MongoDB with role-based access, Confluent Kafka job events and real-time deployment logs over WebSockets.
- For client EXL Services, migrated a legacy data warehouse to Python ETL pipelines on AWS Redshift, orchestrated with Airflow and monitored via CloudWatch.
type a question · answers are pre-written
04
Ask about my work
Answers appear here. Everything runs in your browser — nothing is sent anywhere.
engineering notes · full pieces on medium
Building PseudoChat: an AI-powered Slack chatbot with webhooks, FastAPI and Together AI
Wiring Slack events to an LLM through a verified webhook endpoint — and designing around Slack's three-second deadline.
Managing infrastructure with Pulumi and S3: a journey into code-driven IaC
Why self-managing Pulumi state in a versioned S3 bucket beats the default backends when control and compliance matter.
Exploring Confluent Kafka: a developer's experience in real-time data processing
Tracking job-run status across distributed systems by treating Kafka topics as tables instead of reaching for a database.
My first dive into JWT authentication with FastAPI
What implementing token auth for the first time taught me — about JWT, and about learning unfamiliar things properly.
email · github · linkedin · medium
06
Contact
Let's build
something reliable.
Open to full-stack and AI-engineering roles. Have a project, a question, or just want to talk systems? Reach out.
View resume