What I would measure before shipping an AI workflow
A practical release evaluation that checks outcomes, authority, repeated runs, latency, and cost. Includes a small harness that exposes a misleading improvement.
8 min read
blog.md
Blog
Writing by Subodh Jena on software engineering, LLMs, agent workflows, and the decisions behind building products.
Tool calls that can be retried safely
What a tool should remember when a request succeeds but its response disappears. A runnable SQLite example covers duplicate delivery, conflicting intent, and concurrent retries.
8 min read
A voice assistant has more than one latency problem
Separate turn detection, response generation, playback, and interruption time before trying to make a voice assistant faster. Includes a runnable trace analyzer.
8 min read
Before building an agent, map the workflow
A practical way to find where an agent belongs in a business process, using a workflow map, explicit decision boundaries, and a runnable support example.
10 min read
Graph Engineering Is Not a New Discipline. That Is the Point.
A new label is going around for building agentic systems as explicit execution graphs. The label is barely real. The shift underneath it is, and the evidence for and against it is more interesting than either side admits.
12 min read
I Read the Code
The planner, coder, judge split is a good harness: an expensive model plans, a cheap one writes, the expensive one judges, with the budget spent where judgment lives. But an LLM judge is a filter, not an alibi; it catches the cheap model's misses, not whether the plan solved your problem or whether the edge cases the models quietly agreed on match the ones your users will find. The one step that does not delegate is the last: somebody opens the diff and reads it.
1 min read
Stop Building Agent Demos. Start Building Agentic Software.
Agentic AI is in its demo phase, where a chain of prompts looks magical in a notebook but falls apart in production. The discipline that separates the two is knowing where reasoning belongs to the model and where reliability belongs to ordinary code.
11 min read
What 'AI-Native' Actually Means
A story about a team building an AI-native product. They design the UI first and the model breaks it. They delete the UI and the users break. They hand the agent their old API, then make people learn a control panel. Then they find where the interface actually belongs.
3 min read
The Agentic Future and the Token Economy
The agentic future is now mainstream consensus, and tokens have become the unit of account for the intelligence economy. But tokens are a unit of labor, not money; the real money between agents is stablecoins; and the most actionable move is building software agents can consume.
13 min read
SynthID: How Google Watermarks AI Content
Google has watermarked over 10 billion pieces of AI content with SynthID. The text version is open source, the algorithm is genuinely clever, the detector portal is free for journalists. It is also not a silver bullet, and the academic literature has been clear about why.
14 min read