# adamkalamchi.com

> Personal site of Adam Kalamchi — AI transformation and data infrastructure operator. Writing on AI transformation, data infrastructure, agentic engineering, and the messy industries (real estate, mortgage, insurance, consumer credit) where those tools matter most.

Adam Kalamchi is an AI transformation and data infrastructure operator. He raised $33M from Bessemer Venture Partners, RRE, and MetaProp to build Staircase, a mortgage infrastructure platform with 100+ integrations and an engineering team of 150 built across four continents over six years. He co-founded Elephant Protocol, which tokenized 10.5M Florida properties onchain at 100x cheaper than incumbents. Earlier career: McKinsey, Mubadala (semiconductor portfolio), BCG. Education: Harvard BA Economics cum laude, Columbia MBA. Now interim Chief Data Officer at a Point72/Värde-backed specialty finance platform, and being pulled toward deep tech and the physical sciences.

## Articles
- [A Wrapper with a Loop Is Still a Wrapper](https://adamkalamchi.com/articles/60-agentic-engineering/): Most companies saying "agentic" have a prompt template calling an API. Real agentic engineering is agents that create, QC, and compete with other agents.
- [The Prompt Is the Last 5%](https://adamkalamchi.com/articles/23-context-engineering/): Context is the 95%. The prompt is the last 5%. Why the retrieval, assembly, and layering pipeline determines AI output quality.
- [AI Readiness Scorecard](https://adamkalamchi.com/articles/43-ai-readiness-scorecard/): A pre-qualification framework for AI transformation that measures organizational willingness to change, not technical readiness.
- [3 People. 90 Days. Shipped.](https://adamkalamchi.com/articles/27-pod-model/): Why a 3-person embedded team outperforms larger teams for AI transformation, and the 30/30/30 methodology that makes it work.
- [Enterprise AI: The Organizational Willingness Gap](https://adamkalamchi.com/articles/17-enterprise-ai-transformation/): Why the gap between AI demo and AI production isn't a technology problem -- it's an organizational willingness problem that most companies refuse to confront.
- [Two People Matched My Email Style. Two.](https://adamkalamchi.com/articles/03-systems-thinkers/): Systems thinkers with business background are the highest-leverage people in AI -- and the disadvantage they faced in narrative institutions is now their edge.
- [The Three-Layer AI Workforce](https://adamkalamchi.com/articles/67-three-layer-ai-workforce/): Most enterprise AI deployments build workers — agents that automate recurring processes. The compounding leverage comes from two layers most teams skip.
- [The P&L Test](https://adamkalamchi.com/articles/68-pl-test/): The only honest measure of whether AI transformation worked is whether it shows up in the operating statement. Everything else is theater.
- [A Heartbeat Into Every Part of the Business](https://adamkalamchi.com/articles/69-decision-engine/): How we used AI agents to catalog every measurable thing in a business, build the ontology, and replace a status-report culture with a decision engine that surfaces only what requires action.
- [One Pipeline Pretending to Be Ten Businesses](https://adamkalamchi.com/articles/14-mortgage-default-pipeline/): The mortgage industry is cost-plus. Every intermediary doubles the margin. One pipeline treated as ten businesses costs homeowners $90 billion a year.
- [The Mortgage Value Chain, Part 1: Who Gets Paid and Why](https://adamkalamchi.com/articles/52-value-chain-part-1/): Real estate agents, the MLS, the $31 billion data infrastructure tax, and the 6% commission structure that costs American homeowners $97 billion a year.
- [Mortgage Value Chain, Part 2: The Hidden Rate Tax](https://adamkalamchi.com/articles/53-value-chain-part-2/): How mortgage broker commissions and lender gain-on-sale get buried in your interest rate instead of appearing on your closing statement.
- [The Mortgage Value Chain, Part 3: Servicing Is Just Billing](https://adamkalamchi.com/articles/10-debt-servicing-stack/): In any other industry, this would be called billing. Mortgage invented a $200B industry by renaming billing and wrapping it in compliance jargon.
- [Title Insurance: A $20 Billion Tax on Bad Infrastructure](https://adamkalamchi.com/articles/09-title-insurance/): How a 5% loss ratio industry survives on regulatory capture and fragmented property records
- [The Proptech Pivot: Workflow to Data Company](https://adamkalamchi.com/articles/11-proptech-to-data/): The predictable pattern of proptech companies pivoting from workflow automation to data intelligence -- and how to spot it early.
- [What Changes When AI Remembers You](https://adamkalamchi.com/articles/16-ai-thinking-partner/): The gap between a stateless AI assistant and a persistent thinking partner changes how you make decisions, not just how you get answers.
- [What I Built When RAG Wasn't Enough](https://adamkalamchi.com/articles/39-personal-rag/): Three iterations to build RAG. Then the discovery that retrieval was only one layer of a much larger problem.
- [Delaware Is a $40T Asset Manager Disguised as a State](https://adamkalamchi.com/articles/51-delaware-saas-platform/): The first state to operate like a SaaS platform already exists. It just doesn't know it yet.
- [The Cost-Plus Industry: A Mortgage Retrospective](https://adamkalamchi.com/articles/38-staircase-retrospective/): Two million people make their living off the inefficiency. We just didn't think change was possible.
- [Notes on Being a Software CEO](https://adamkalamchi.com/articles/65-being-a-software-ceo/): Operating notes from six years as a software CEO -- people, architecture, customers, and the sequencing discipline that encodes founder judgment.
- [Why Decentralized Protocols Beat Startups](https://adamkalamchi.com/articles/64-why-decentralized-protocols/): Why decentralized protocols solve problems that neither startups nor corporations can -- and why the advantage compounds.
- [Why NVIDIA Won the AI Era (and a Sovereign Fund Didn't)](https://adamkalamchi.com/articles/30-inside-sovereign-wealth/): Three years inside a sovereign wealth fund during the GlobalFoundries acquisition, and what the semiconductor bet teaches about buying capability, not just raw AI, in the compute era.

## Papers
- [Bitstrapping](https://adamkalamchi.com/bitstrapping/): Bitcoin-Native Funding for Decentralized Protocols
- [Elephant Protocol](https://adamkalamchi.com/elephant-protocol/): Decentralized Infrastructure for Real Estate Truth

## Pages
- [About](https://adamkalamchi.com/about/): Full biography and career timeline
- [Stats](https://adamkalamchi.com/stats/): Career highlights and metrics
- [Articles](https://adamkalamchi.com/articles/): All articles on AI transformation, real estate, and systems thinking
- [Code](https://adamkalamchi.com/code/): Open source projects including Elephant Protocol
- [Posts](https://adamkalamchi.com/posts/): LinkedIn posts — AI, agentic engineering, real estate, and software building

## Recent LinkedIn Posts
- [2026-06-18](https://www.linkedin.com/feed/update/urn:li:activity:7473442722801156098): For CEOs struggling with where to start with AI adoption, I have good news, bad news, and a suggestion. First, the good news: Everyone is feeling extreme FOMO compared to everyone else who seem to be 
- [2026-06-12](https://www.linkedin.com/feed/update/urn:li:activity:7471176718691721216): Confirmed, claude code ultracode will crush tokens.
- [2026-06-09](https://www.linkedin.com/feed/update/urn:li:activity:7470086564451352576): I hit a breaking point. 5+ terminals open with the same title. scrolling up to remember which was which. Getting lost in gits, commits, stashes, stomps, pushes, pulls, deploys. So I created a message 
- [2026-06-07](https://www.linkedin.com/feed/update/urn:li:activity:7467934163883622400): Real estate and mortgage behave more like defense than finance.  The incentives all point the same way - higher prices: - Real estate and mortgage are 'ad valorem' industries - 'Ad valorem' has the sa
- [2026-05-31](https://www.linkedin.com/feed/update/urn:li:activity:7465841539009736705): I made a big change to my workflow recently - added the concept of a "workbench". A workbench is a deep dive into a topic, company, person, market, etc. where I want to do deep agentic research, aggre
- [2026-05-15](https://www.linkedin.com/feed/update/urn:li:activity:7461132162172530688): I took things to their logical conclusion (late) last night ... made an Adam agent that acts as the reviewer at each stage-gate in the build process. Got tired of saying the same things over and over 
- [2026-05-15](https://www.linkedin.com/feed/update/urn:li:activity:7460667277839953920): Couple things I've done recently which have been helpful: 1/ I created a sitemap of the product I'm building. Coding agents start each session by reading the sitemap, then do the work, then update the
- [2026-05-15](https://www.linkedin.com/feed/update/urn:li:activity:7454551767037595648): I've had a few friends reach out to me to talk shop about AI and how to transform their orgs. Is AI coach a thing? Regardless, if you're thinking about how to kick-start AI use within your org, feel f
- [2026-05-15](https://www.linkedin.com/feed/update/urn:li:activity:7453408365185196032): Couple helpful notes I've found for building with agents: 1/ I created a team with 5 personas: orchestrator, codebase expert, researcher, builder, qa agent. Each is assigned a notable public figure as
- [2026-04-15](https://www.linkedin.com/feed/update/urn:li:activity:7455604455749763072): We’re not hiring traditional software engineers anymore. We’re hiring Architectural Operators — people who design systems that build software using AI agents, rather than writing code themselves. We’r
- [2026-04-15](https://www.linkedin.com/feed/update/urn:li:activity:7446943811454451712): I've been having a blast with Claude Code. I'm building a few things right now, but here's one that's probably relevant to most people — I call it Dojo. A personal AI operating system. Six apps sharin
- [2026-04-15](https://www.linkedin.com/feed/update/urn:li:activity:7442664955423637504): State-of-the-art frontier AI models are <<$10 per hour if you use them heavily non-stop for the full hour ... That's insane for having a genius side-kick. You aren't using them enough!
- [2026-04-06](https://www.linkedin.com/feed/update/urn:li:share:7446943808719785985): I've been having a blast with Claude Code. I'm building a few things right now, but here's one that's probably relevant to most people — I call it Dojo. A personal AI operating system. Six apps sharin
- [2026-03-26](https://www.linkedin.com/feed/update/urn:li:share:7442664954907947008): State-of-the-art frontier AI models are <<$10 per hour if you use them heavily non-stop for the full hour ... That's insane for having a genius side-kick. You aren't using them enough!
- [2026-03-13](https://www.linkedin.com/feed/update/urn:li:share:7438202720903278592): Using Claude code feels like being a kid again - magic. I say that without hyperbole. If you haven't tried it, you really need to. For a non-technical person, it'll be the easiest way to get a gut-lev
- [2025-11-12](https://www.linkedin.com/feed/update/urn:li:share:7394131215022460929): The 50-year mortgage discussion misses the point entirely. It's money printing by another name. I recently walked a family member through mortgage optimization. Running just 4 basic scenarios across a
- [2025-11-05](https://www.linkedin.com/feed/update/urn:li:share:7391847405341585408): Compromise kills the protocol flywheel. Protocols are counterpartyless, code-native, eternal. VCs need counterparties, documents, and exits. A structural mismatch runs deep between how protocols opera
- [2025-10-29](https://www.linkedin.com/feed/update/urn:li:share:7389040746546270208): Protocols are going to eat startups from the bottom, exactly like startups did to corporations. First solve problems even startups can't economically justify, then move upmarket.  Building Elephant Pr
- [2025-10-22](https://www.linkedin.com/feed/update/urn:li:share:7386509730514587648): Prediction markets would predict fire risk better than insurance premiums. Participants would get granular—roof condition, lot slope, wind direction, distance for fire hydrant, etc. But insurance prem
- [2025-10-15](https://www.linkedin.com/feed/update/urn:li:share:7384205282517721088): A major P&C insurer tested Elephant Protocol data on their hardest underwriting problem: roof age in disaster zones. One county in Florida revealed $67M in preventable Hurricane Ian losses. https://ln

## Optional
- [llms-full.txt](https://adamkalamchi.com/llms-full.txt): Concatenated full markdown of every article
- [Sitemap](https://adamkalamchi.com/sitemap.xml): All indexable URLs with lastmod