Services
AI agents, platforms, cloud and DevOps, designed and shipped from idea to production, on infrastructure designed for 1M+ users.
Tool-calling AI agents, LLM apps and RAG chatbots that act on your real business data, with guardrails, escalation paths and per-request access checks.
Proof: Built a clinical AI agent that pulls live patient records over MCP, and MCP + RAG tools that automate workflows at Implevista.
See the workModel Context Protocol servers and AI gateways that connect OpenAI, Claude and Gemini to your systems behind one secure API.
Proof: Built Mediflow's AI gateway (OpenAI, Claude, Gemini) and its MCP server for real patient data.
See the workMulti-tenant SaaS platforms and internal developer platforms: identity, billing, dashboards, APIs, SDKs and CLIs that other teams build on.
Proof: Architected a multi-tenant travel SaaS platform, and built RivoCloud with Stripe billing, REST API, SDK and CLI.
See the workProduction infrastructure on AWS, GCP or Azure: containers, load balancing, CDN and object storage, sized for cost and latency.
Proof: AWS ECS/EC2/ALB/CloudFront, a GCP serverless pipeline that cut processing time 70%, and caching designed for 1M+ users.
See the workPipelines that test, scan and ship on every merge: Docker, Kubernetes, GitHub Actions, ArgoCD GitOps, blue/green deploys and observability.
Proof: RivoCloud ships via ArgoCD GitOps with health-gated blue/green cutover; multi-stage Docker pipelines at Implevista.
See the workEmbed with your team or your customer's team, understand the real problem, and ship a working system into their production environment.
Proof: Integrated GDS, aggregator and SOAP APIs into one booking engine, led a cross-functional team of 10+ engineers, and shipped AI onto live data.
See the workTake an AI or SaaS product from idea to production: architecture, backend, frontend, CI/CD and launch, built to scale from day one.
Proof: Designed and shipped Mediflow and RivoCloud end-to-end.
See the workTelemedicine platforms, AI health assistants, clinical AI agents, prescription OCR and digital health records.
Proof: Built Mediflow: video consultations, the Medora AI assistant and a clinical AI agent for doctors.
See the workHealthcare & wellness
Telemedicine, AI health assistants, clinical AI agents
Media & OTT
AI-powered content platforms, recommendation engines
Travel & B2B SaaS
Multi-tenant platforms, booking engines, CRM and CMS
Cloud & developer tools
CDN, object storage, Kubernetes, APIs, SDKs and CLIs
Discover
Pin down the user problem, the data you have, and what success looks like.
Architect
Choose models, boundaries and infrastructure before writing code.
Build
Ship in small, working increments with CI/CD from the first week.
Launch & scale
Deploy, monitor, and tune cost and latency as real traffic arrives.
AI agents, LLM apps and RAG pipelines, MCP servers and multi-LLM integrations, plus the platform, cloud and DevOps infrastructure they run on: multi-tenant SaaS, APIs and SDKs, Kubernetes, CI/CD and observability.
Yes. Mediflow and RivoCloud were both designed and built end-to-end, from architecture through CI/CD to production deployment, and the same approach is used on AI products at Implevista.
OpenAI, Anthropic Claude, Google Gemini and OpenRouter, with the Model Context Protocol (MCP) for tool-calling and Retrieval-Augmented Generation (RAG) for grounding, on Node.js, TypeScript, Go, Python and AWS, GCP or Azure.
Yes. The work so far has been end-to-end delivery: integrating GDS, aggregator and SOAP APIs into one booking engine, leading a cross-functional team of 10+ engineers on a multi-tenant SaaS platform, and shipping AI agents onto live production data.
Yes. AWS (ECS, EC2, ALB, S3, CloudFront, Route 53), GCP serverless and Azure, with Docker, Kubernetes (k3s), GitHub Actions, ArgoCD GitOps, blue/green deploys and Prometheus, Grafana and Loki for observability.
No. Healthcare is one industry (Mediflow). Other work spans an AI-powered OTT media platform with a recommendation engine, a multi-tenant travel SaaS platform, and RivoCloud, a CDN and cloud storage product.
Book a call at https://cal.com/dihanrh or email dihanrh@gmail.com with a short description of the product and your timeline.
Have an AI product in mind?
Tell me what you're building and I'll reply within 24 hours.