adwibha@MacBook:~
$ whoami
name : Akhil Dwibhashyam
location : Austin, TX · Open to remote
stack : AWS · Terraform · Kubernetes · Python · FastAPI · TypeScript/Node
focus:
backend : REST APIs, FastAPI/Node, service design, PostgreSQL
infrastructure : IaC, ECS/EKS, CI/CD, zero-downtime deployments
reliability : SLIs/SLOs, observability, incident response automation
ai-systems : Multi-agent systems, LLM cost optimization, MCP servers, agentic workflows
certifications:
- AWS Solutions Architect – Associate
- HashiCorp Certified Terraform Associate
- Certified DevOps Engineer (CDE)
- McKinsey Forward Program
- Engineer AI Agents with Agent Development Kit (ADK)
education : Master of Science in Computer Science
currently : Software Engineering · DevOps · SRE · AI Infrastructure roles5+ years building backend services, cloud infrastructure (AWS, Azure, GCP), and DevOps pipelines. MS in Computer Science.
Increasingly focused on production agentic AI: multi-agent orchestration with LangGraph, LLM cost optimization, and observable autonomous systems. Recent work spans production research agents, semantic caching (60% LLM cost reduction), and real-time analytics platforms.
Before AI, I ran large-scale cloud migrations and built EKS deployment pipelines. Core intersection: infrastructure reliability meets agentic AI.
|
Cloud Infrastructure Production AWS environments with Terraform IaC, ECS/EKS orchestration, VPC networking, and automated CI/CD pipelines. |
Agentic AI & Backend Multi-agent systems with LangGraph orchestration, REST APIs (FastAPI, Node), LLM cost optimization, MCP servers, and streaming architectures. |
Reliability & Observability Observable agentic systems, SLI/SLO frameworks, incident response automation, and production monitoring stacks (Prometheus, Grafana, Splunk). |
|
Autonomous agentic system for research synthesis and iterative refinement Production-grade multi-agent orchestration with LangGraph. Autonomous research synthesis, iterative refinement, and tool orchestration. Real-time API integration, context management, and agentic workflows at scale. |
Enterprise analytics with LLM-powered natural language queries
Semantic caching reduced LLM API costs by 60%. Real-time PostgreSQL KPIs, D3.js visualization, token tracking, rate limiting. 99.9% uptime. |
|
Containerized deployment pipeline for multi-module enterprise application
Containerized multi-service app and migrated to EKS. Built Jenkins CI/CD pipelines with automated rollouts and zero-downtime deployments. Reduced deployment time by 70%. |
On-premises to cloud storage migration for a Canadian retailer
Migrated Hitachi HUSVM/HNAS storage to AWS. Maintained 99.9% uptime throughout migration with zero data loss. Designed HA/DR architecture and runbooks. |
|
10-module learning curriculum for Google Agent Development Kit Hands-on learning resource. Agent configuration, tool orchestration, multi-agent systems, LangGraph integration, and production patterns. |
Google Agent Development Kit — TypeScript Contributor to official Google ADK. Building agentic capabilities into the TypeScript ecosystem. |
|
LLM cost optimization via intelligent model routing Routes prompts to the cheapest capable model. Per-request cost telemetry, OpenAI-compatible API, multi-provider support. |
Observability stack for agentic systems 6-service stack with SLI/SLO alerting. One |
|
Document ingestion → semantic search → local LLM response pgvector RAG with Redis caching (5ms latency on repeated queries). Ollama local inference, MCP server integration. |
Agentic search with iterative query refinement Agent-driven RAG. Autonomous query refinement, multi-step search strategies, and relevance feedback loops. |
- How to Run Open-Source LLMs Locally with Ollama — The Fastest Way to Start
- Monitoring AWS EKS with Prometheus and Grafana
- Infrastructure as Code (IaC) with Terraform on AWS — The What, Why & How
View all on Medium. Longer posts and notes also live on blog.adwibha.me.
