CRM
A CRM SaaS running the full lead-to-conversion lifecycle for education consultancies and visa institutes.
- RBAC / ABAC
- Realtime
- Kubernetes
A Dipesh Chaulagain production · click to skip

Not just using AI — engineering resilient, observable and scalable agentic systems that hold up in production.
Everyone is using AI. Few are engineering it. I build agentic systems that fail gracefully, explain themselves and scale without drama — software where the model is a component, not a miracle.
1callTool.pipe(2 Effect.retry(Schedule.exponential("250 millis")),3 Effect.timeout("8 seconds"),4 Effect.orElse(() => fallbackProvider)5)
Fullstack products shipped end-to-end — from database schema to the pixel — across industries where the stakes are real.
A CRM SaaS running the full lead-to-conversion lifecycle for education consultancies and visa institutes.
A B2B logistics platform for enterprise fleets, where trip schedules are anything but predictable.
Clinical data pipelines where integrity isn't a feature — it's the requirement.
One control plane for configuring and querying SIEM deployments across cloud and on-prem.
A multi-tenant, AI-native platform for creating, managing and automating social content across every network.
Taking everything learned across these worlds and rebuilding it around agents that are reliable enough to trust with real work.

Interactive, deeply-researched articles on building AI-native features with modern tools — live code, playable diagrams and real traces. High signal, zero fluff.
Typed failures, retries, timeouts, fallbacks and crash-proof steps for agents that never silently die.
Server functions, AI SDK streams and components that render while the model thinks.
Spans, token budgets and evals wired into the same telemetry as the rest of your stack.
Supervision trees as a mental model for fault-tolerant, multi-agent systems.