Kam AI Framework9 min read
Kam AI Is Becoming an Operating System
May 24, 2026
Why Kam is moving from AI chat into a production loop of traces, labels, graders, fixtures, release gates, and agentic work.
Read noteKam AI
Blog
Field notes on Market Shape, sportsbook odds, prediction markets, freshness, missing info, saved reads, and better reads before you bet.
Field notes
Catch moves. Don’t chase.
Field notes and prompt libraries on Market Shape, what moved, why it matters, what is missing, and what to verify next.
16 posts
New to Kam
StartStart here. Learn why Kam reads sportsbooks, prediction markets, and saved context before you bet.
Read pathAsk better questions
Use prompts that give Kam the game, sportsbook, prediction market, missing info, and next check.
Read pathTrust the card
See how Kam checks freshness, missing info, and answer quality before you trust the read.
Read pathHow Kam works
Understand how sportsbook context, prediction markets, Ask Kam, saved reads, and review connect.
Read path
System Design8 min read
The Kam Intelligence Registry
May 24, 2026
How a workload-centered registry connects traces, labels, fixtures, judge evidence, agentic runs, work items, and release gates.
Read noteTelemetry7 min read
Production Traces Need Workload IDs
May 24, 2026
Why Kam treats workload IDs as the unit for trace slicing, deduplication, dataset creation, and release scorecards.
Read noteKam Evals8 min read
Deterministic Graders Before LLM Judges
May 24, 2026
Why Kam checks route, entity, source, freshness, denominator, and contract facts before asking an LLM judge whether an answer was useful.
Read noteKamOps8 min read
Human Labeling Is the Center of KamOps
May 24, 2026
How KamOps turns failed traces into approved labels, fixtures, and review packets without letting automation become the source of truth.
Read noteRelease Quality8 min read
Fixtures Turn AI Lessons Into Release Gates
May 24, 2026
How Kam promotes approved labels into fixtures and uses workload gates to prevent the same answer failures from shipping again.
Read noteAgentic Systems8 min read
KamAgentic Is for Bounded Internal Work
May 24, 2026
Why normal chat stays route-contract-first while KamAgentic handles trace labeling, fixture promotion, release packets, and long-running ops.
Read noteKamSRE8 min read
KamSRE Turns AI Quality Into Scorecards
May 24, 2026
How daily health digests, workload scorecards, drift checks, and release evidence make Kam AI operable.
Read noteArchitecture9 min read
Use Open Source for Plumbing, Keep Kam Intelligence Custom
May 24, 2026
How Langfuse, Phoenix, OpenTelemetry, Label Studio, Argilla, promptfoo, DeepEval, Evidently, Temporal, Dagster, and LiteLLM fit around Kam.
Read noteLessons Learned10 min read
Lessons From Building the Better Kam Framework
May 24, 2026
What Kam learned moving toward a framework of KamSRE, KamOps, KamEvals, KamAgentic, labels, fixtures, and workload scorecards.
Read noteMarket Shape16 min read
Catch Moves. Don’t Chase.
May 8, 2026
Learn why Kam watches sportsbooks, prediction markets, and saved context before it answers a market question.
Read noteTrust20 min read
How Kam Checks an Answer Before You Trust It
May 9, 2026
See how Kam checks freshness, missing info, scope, and answer quality before you trust a move card.
Read notePrompt library14 min read
Sports Questions That Get a Better Read
May 9, 2026
Use practical prompt upgrades for market scans, line moves, prediction-market context, missing info, watchlists, and postgame review.
Read noteHow it works10 min read
How Kam AI Is Built
May 7, 2026
Understand how Kam turns sportsbook odds, prediction markets, schedules, saved reads, and review into one product loop.
Read noteMarket Shape6 min read
Know the Board Before You Fire
May 7, 2026
Learn how Kam slows down market chasing by showing what moved, what is missing, and when to pass.
Read noteProduct thesis5 min read
Use Kam as a Second Opinion
May 6, 2026
Learn how Kam speeds up market checks and review while keeping the final bet human.
Read note