The AI Runtime

Home
RSVP for Boston Events
Lessons From the Trenches
How It Works
Tools & Workflows
Career & Industry
Builder Spotlight
Archive
About
A Portfolio That Practices MRE
Vishnu Purohitham’s four shipped projects are a worked example of Model Reliability Engineering — and a soft hit on most of the AIfolio.
 • The AI Runtime
Three Weeks of Opus 4.7 in Production: What Teams Are Actually Reporting
The launch numbers were one story. The production patterns are a different one.
 • The AI Runtime
Inside Mintlify’s Agent Stack
A teardown of the two-harness architecture — async sandboxes for writes, virtual filesystems for reads — and what it teaches about wrapping a model in…
 • The AI Runtime
How Vertical Agents Self-Improve in Production
Field notes on the harness loop at Harvey, Hippocratic, Anterior, and Azure SRE — where production failures compound into skill without retraining the…
 • The AI Runtime
Felix Is a Harness, Not a Model: How Rogo Built an Agent for High Finance
Rogo just raised $160M Series D led by Kleiner Perkins. The architecture behind their Felix agent is what AI engineers should be studying.
 • The AI Runtime

April 2026

Privacy Filter Is Not an LLM
OpenAI’s open-weight PII model is a bidirectional token classifier — what that architecture buys, where the headline benchmark misleads, and why…
 • The AI Runtime
Shadow AI Agents
Your enterprise has more AI agents than employees. Most don’t have identities, owners, or audit trails. Agent identity is the reliability surface that…
 • The AI Runtime
Builder Spotlight - Armaan Agrawal ships like a forward-deployed engineer already
Seven production systems, one repeatable pattern, and the real-world skills most new grads don’t show up with.
 • The AI Runtime
The Vercel Breach RCA: Agent Identity Is the New Attack Surface
One OAuth grant, one compromised AI vendor, one platform breach. Every team deploying agents shares the same architecture.
 • The AI Runtime
OpenAI’s AI Deployment Playbook Is Missing a Chapter
Their whitepaper nails the org chart. It ignores the engineering discipline that determines whether AI products actually stay in production.
 • The AI Runtime
The Eval Lifecycle: What Actually Happens Between “Proof of Concept” and “Production”
Most AI projects die in the gap between “it works on my laptop” and “it works in production.” The eval lifecycle is the bridge nobody teaches you to…
 • The AI Runtime
Your AI Strategy Doesn’t Need More Use Cases. It Needs a Production System.
Why most enterprise AI strategies fail at the same point — and the five decisions that separate companies shipping AI products from companies running…
 • The AI Runtime
© 2026 Kranthi Manchikanti · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for great culture