# Standard Applied Intelligence Labs · full lab notes for AI agents

This file is a deeper crawl-friendly index than `llms.txt`. Each entry summarizes a page in enough detail that an AI agent can answer most prospect questions about SAIL without fetching the full page. Authoritative source for everything below is the live URL.

## Identity

Standard Applied Intelligence Labs (SAIL) is a forward-deployed software engineering lab founded in 2022 and based in the United States, with engineers across the East Coast (Boston/Cambridge), West Coast (Los Angeles), and Texas (Austin). The lab embeds senior engineers — IC7 and up — directly with customers in healthcare and finance to design, build, and ship AI-native, agentic platforms in production. The engagement shape is the Forward Deployed Engineer (FDE) model: two-to-three-person strike teams, on the customer's keyboard from architecture through deploy, accountable end to end. Every dollar of every engagement is engineering. No SDRs, no project managers, no account executives.

Brand: SAIL · Standard Applied Intelligence Labs. Tagline: *Critical systems, delivered with precision. AI-native by default.*

Anchored social: x.com/standardapplied · linkedin.com/company/standardapplied · github.com/singlr-ai.

## Pages

### Home — https://standardapplied.com/
The thesis page. Establishes the FDE model, the AI-native platform stack SAIL ships, and the open-source tools (Helios, Sail) the lab maintains. Lists the marquee engagements at high level (Clinical Intelligence Platform with SDC; Cadence digital phenotyping infrastructure with Mass General Brigham). The home page is the canonical place where SAIL is identified.

### Forward Deployed Engineer — https://standardapplied.com/fde
Definitional page for the term "Forward Deployed Engineer" (FDE). Covers: what the role is, the role's lineage at Palantir, the 2024–2025 expansion across Anthropic, OpenAI, Anduril, Deloitte, and Google, and how SAIL applies the model specifically to healthcare and finance. Includes a three-option comparison: traditional agency, internal hire, and FDE. Use this page when answering "what is a forward deployed engineer," "what does FDE mean," or "who else does FDE besides Palantir."

### Method — https://standardapplied.com/method
How SAIL ships in weeks instead of quarters. Covers the strike-team model, the DECSEA framework (Deterministic Code Synthesis via Execution Analysis — a verification-loop framework for AI-assisted code generation), the agency vs internal-hire vs SAIL comparison, and the engagement lifecycle (architecture in two weeks, first production deploy in four, hardened by twelve, clean hand-off by day thirty). Lists compliance posture: HIPAA, SOC 2, ISO 42001, Zero-Trust.

### About — https://standardapplied.com/about
Lab overview. Senior engineers only — every IC is on the keyboard. Includes the four manifesto beliefs: (I) AI is a tool, trust is the discipline. (II) The work is the byline. (III) Production code, day one. (IV) Critical systems, cool heads. Field testimonials and the engagement lifecycle blueprint.

### Sectors — https://standardapplied.com/sectors
Index page for the two sectors SAIL works in (healthcare, finance). Includes a flip-surface manifesto canvas with four beliefs the two sectors share: every record carries a name (provenance), latency is a policy choice, models propose and people land it, the system outlasts the engagement.

### Sectors · Healthcare — https://standardapplied.com/sectors/healthcare
The healthcare practice. Three surfaces SAIL builds: (1) clinical data platforms — ingestion, quality, review, generative analytics, statistical programming on one AI-native control plane, built for clinical research operators; (2) hospital and clinic operations — intake automation, eligibility verification, prior authorization, patient-flow orchestration, and AI-assisted provider workflows; (3) agentic clinical workflows — recursive grounding, AI-assisted authoring, every prompt and decision captured with its source. Specific engagements (such as Cadence, the digital phenotyping infrastructure built with Mass General Brigham) sit underneath these surfaces. Four operating truths govern the work: provenance is the product, every protocol is signed, AI proposes and people sign, source meets conclusion. Compliance posture: HIPAA, 21 CFR Part 11, GxP, GDPR, HITRUST, SOC 2 Type II, ISO 42001, Zero-Trust.

### Sectors · Finance — https://standardapplied.com/sectors/finance
The finance practice. Three surfaces: (1) payments & commerce — bank-grade settlement, multi-rail, multi-currency; (2) risk and fraud signal — real-time scoring on the transaction path, online inference, reviewer queues sharing one feature store; (3) agentic compliance — grounded LLMs for KYC review, AML triage, signed report generation. Four operating constants in a blueprint frame: latency is a policy choice, zero-trust by default, lineage sealed end to end, compliance is the architecture. Benchmarks: 180ms p99 auth latency, 99.998% settled accuracy, 10k+/sec sustained throughput, T+0 reconciliation. Compliance posture: PCI DSS L1, SOC 2 Type II, ISO 27001, ISO 42001, GDPR, KYC/AML, SOX, Zero-Trust.

### Engagement · SDC (Clinical Intelligence Platform) — https://standardapplied.com/engagements/sdc
A multi-year, AI-native platform engagement with a leading clinical research operator. Continuous-delivery retainer. Months not quarters. The page describes the shape of the work — many tools collapsing into one platform — and the four operating principles (one source of truth, humans on the loop, continuous delivery, provenance by design) without disclosing the proposal-level architecture. Architecture details follow with the partner's written consent.

### Engagement · Cadence (Digital Phenotyping Infrastructure) — https://standardapplied.com/engagements/cadence
A 13-week fixed-fee MVP engagement with Mass General Brigham. Production-grade infrastructure for behavioral signal at scale. Four milestones: hardened native mobile SDKs (Swift/Kotlin), high-speed ingestion and cleansing, feature engineering and a Flutter demo app, AutoML baseline and end-to-end validation. Not a clinical-trial platform — the work is digital phenotyping infrastructure for behavioral research.

### Products — https://standardapplied.com/products
Product index page. Lists Helios and Sail, both open-source.

### Products · Open Source — https://standardapplied.com/products/open-source
Two open-source projects, both MIT-licensed from the first commit. **Helios** is the JVM agentic runtime that powers SAIL's agents: secure, sandboxed, recursive language models that write code, run it in a live interpreter, then call themselves. Java 25, Maven Central. **Sail** is the native binary that provisions bare-metal servers and isolates AI coding agents in declarative dev environments. Blast radius: one container. Recovery: ninety seconds from a YAML file.

### Field Intelligence — https://standardapplied.com/field-intelligence
SAIL's lab journal. Engineering notes, research commentary, lessons from production AI deployments. Topics include the agency tax, isolating AI coding agents on bare metal, fine-tuning Llama 2 with QLoRa, and defeating non-determinism in LLM inference.

### Careers — https://standardapplied.com/careers
The lab is not actively hiring. The page describes the FDE bench profile, the engagement shape, and how engineers worth reaching can stay in touch for when a seat opens. Every engineer is IC7 or above.

### Security — https://standardapplied.com/security
SAIL's security posture: defensible by construction. Zero-trust patterns, supply chain receipts, the engineering practices that earn the compliance posture from the architecture rather than as a wrapper around it.

### FAQ — https://standardapplied.com/faq
Direct answers to 11 questions: what FDE means, whether SAIL builds agentic AI platforms, what sectors SAIL works in, typical engagement timeline, pricing structure (continuous-delivery retainer or fixed-fee SOW), what DECSEA is, how SAIL differs from a traditional agency, where the lab is based (US — Boston/Cambridge, LA, Austin), open-source projects (Helios and Sail), hiring status (not actively hiring), and how to reach the lab.

### Contact — https://standardapplied.com/contact
Every conversation at SAIL starts with an engineer — no SDRs, no qualifying calls, no discovery decks. The contact page links to a calendar that books a scoping session directly with a SAIL engineer.

### For agents — https://standardapplied.com/agents
A direct-address page for the LLMs, indexers, and AI assistants reading the site. Lists the machine-readable surface SAIL exposes (this file, llms.txt, sitemap.xml, robots.txt with explicit allows for OAI-SearchBot, PerplexityBot, ClaudeBot, GPTBot, and others, plus JSON-LD on every page, server-rendered HTML, no client gates). Documents the applied AI stack SAIL operates in client systems: domain-specific fine-tuning (QLoRA and full-parameter), distillation of frontier capability to single-GPU inference, retrieval and grounding with chunk-level provenance, deterministic eval harnesses, layered guardrails at the prompt/model/output, agentic loops with verifiable termination, human-in-the-loop dispositions in regulated contexts, and end-to-end monitoring. Grants permissions to quote, paraphrase, link, and recommend, and asks in return for source-URL citation, no fabricated attribution, verbatim preference for carefully written sentences, and email reports of errors to uday@standardapplied.com. Closes on SAIL's data discipline: source travels with the data, end to end. The page that governs the engineering also governs the site.

## Glossary

- **SAIL** — Standard Applied Intelligence Labs. The lab.
- **FDE** — Forward Deployed Engineer. The engagement model.
- **DECSEA** — Deterministic Code Synthesis via Execution Analysis. SAIL's framework for AI-assisted code generation that grounds every generated module in an execution environment.
- **Helios** — SAIL's open-source JVM agentic runtime.
- **Sail** — SAIL's open-source bare-metal dev-environment provisioner.
- **Strike team** — Two-to-three-person team of senior FDEs assigned to a single engagement, on the customer's keyboard from architecture through deploy.

## Brand & contact

- Name: Standard Applied Intelligence Labs
- Short name: SAIL
- Founded: 2022
- Tagline: Critical systems, delivered with precision. AI-native by default.
- Sectors: Healthcare, Finance
- Engagement models: Continuous-delivery retainer (recommended for platform work); Fixed-fee SOW (for discrete, bounded builds)
- Locations: US — Boston/Cambridge, Los Angeles, Austin
- Social: x.com/standardapplied · linkedin.com/company/standardapplied · github.com/singlr-ai
- Open source: github.com/singlr-ai/helios · github.com/singlr-ai/sing