Lakeshore Federal Services LLC

Agentic AI Governance · Treasury Modernization · M-26-04 Disclosure Pack Drafted

Capabilities STATEMENT

UEI GB92MRKPHCE9

CAGE Code 1Z9H3

Business Size Small Business

SAM Status Active

Company Overview

Lakeshore Federal Services LLC is a New Hampshire-based small business focused on AI-augmented modernization for the U.S. Department of the Treasury and federal financial systems. We operate Oceum (oceum.ai), a federal-deployable AI runtime with governance and credential security built into the architecture — production for ~12 months, multi-tenant, end-to-end audit-logged, with zero-knowledge credential mediation and a Governed Execution Engine. Agentic capability runs inside the governance layer, not bolted onto it. Deployable to AWS GovCloud, Azure Government, or fully on-prem / air-gapped with locally-hosted LLMs.

Federal Vertical Focus — Treasury & Federal Financial Systems

We focus our federal delivery capacity on agencies advancing tax administration, federal financial management, and regulatory technology, including:

IRS — Tax modernization, Direct File, case management, customer service AI

Bureau of the Fiscal Service — Payment systems, financial reporting

OCC — Bank examination tech, regulatory analytics

FinCEN — Anti-money laundering, transaction analytics

Departmental Offices — Cross-Treasury enterprise IT

Adjacent agencies — HHS/CMS, GSA, financial-regulatory work

New Hampshire State Focus — Home-State Small Business

As a NH small business headquartered in Derry, we also serve State of New Hampshire agencies with home-state responsiveness, competitive labor rates, and the same AI-augmented modernization approach. Primary state partners include:

NH DoIT — Statewide IT modernization & software development

Revenue Administration — Tax administration tech (federal vertical alignment)

Banking & Insurance Departments — Regulatory analytics & data systems

DHHS — Health & human services modernization

Department of Transportation — Mobility & automation systems

Treasury & Liquor Commission — Financial systems & data

Core Capabilities

Oceum platform — production governed AI runtime with zero-knowledge credential mediation, Governed Execution Engine, AES-256-GCM credential vault, end-to-end audit logging, and 28 OAuth integrations. Customer-deployable to AWS GovCloud, Azure Government, or fully on-prem / air-gapped with locally-hosted LLMs. Powers all AI delivery below.

Governed AI runtime — zero-knowledge credential mediation (PAM-for-AI), policy-enforced execution, audit logging

OMB M-25-22 / M-26-04 readiness — full disclosure pack (AUP, Model/System/Data Cards, feedback channel)

RMF / continuous-ATO automation — built on the governed runtime

Agentic workflows — document triage, case routing, structured extraction over unstructured documents

Modernization & application development — custom software, web apps, APIs, legacy modernization

Cloud architecture & migration — AWS, Azure GovCloud, on-prem / air-gapped

Data analytics & BI — fraud signals, regulatory analytics, dashboards, business intelligence

Document processing & digitization — records digitization, data migration, intelligent extraction

Database design & migration — schema design, ETL, data quality

Cybersecurity advisory & FISMA support — Zero Trust (M-22-09) alignment, ATO-package support

Requirements analysis & feasibility studies — sources-sought support, capability assessments

Technical writing & training development — federal financial & tax administration scopes

NAICS Codes

Code	Description
541511 *	Custom Computer Programming Services
541512	Computer Systems Design Services
541519	Other Computer Related Services
541611	Administrative Management Consulting
541990	Other Professional, Scientific & Technical
561410	Document Preparation Services
611430	Professional & Management Training

* Primary NAICS

Product Service Codes

Code	Description
DA01	Application/App Development Support Services
DD01	Service Delivery Support: ITSM, PM, OpsCenter
DF01	IT Management Support Services

Differentiators — Architectural, Not Aspirational

Zero-Knowledge Credential Mediation — PAM for AI Agents

Agents act on customer-supplied credentials they cannot read. Credentials are AES-256-GCM-encrypted with per-organization key derivation, attached at the network boundary, and never enter agent prompt context. Every credential use and mediated action is audit-logged. This is an architectural property, not a best-effort attestation — and it directly maps to OMB M-22-09 Zero Trust, M-25-22 §3a (no training on agency data), and M-26-04 LLM transparency. Most small AI firms cannot make the same claim.

Governed Execution Engine

Every agent action passes through policy enforcement before it reaches the network. Per-action audit records, scope-bounded execution, and human-in-the-loop checkpoints are first-class primitives, not retrofitted controls. The same engine has been in production for ~12 months across multi-tenant operation.

Customer-Deployable to Federal Environments

Cloud-native, with deployment paths to AWS GovCloud, Azure Government, or fully on-prem / air-gapped. The LLM router supports commercial providers or locally-hosted / in-network models, so closed-environment and sovereign-data deployments do not require external LLM connectivity. NH-based engineering delivery; U.S.-LLC, U.S.-citizen-owned, aligned to M-25-22 §3c “Buy American” emphasis.

OMB M-25-22 & M-26-04 Disclosure Pack — Drafted v0.1

Acceptable Use Policy, Model Cards (per LLM), System Card, Data Card, end-user resources, and feedback channel are drafted v0.1 and available on request. Internal and legal review are pending ahead of v1.0 (target June 30, 2026). M-25-22 vendor-lock-in protections — knowledge transfer, data and model portability, licensing transparency, no-training-on-agency-data — are baked into our standard delivery framework. Most small WOSBs have not addressed these requirements; we have draft artifacts available now.

Engineered Operating Model — Capacity, Not Headcount Overhead

Lakeshore is intentionally small-headcount because the platform does the work. Capacity that conventional firms would staff with billable analysts — automation, document processing, case orchestration, integration scope — is delivered by the runtime instead. The result is lower overhead, faster delivery cycles, and procurement-line cost pressure that scales with software rather than bench size. Continuity risk is addressed structurally: Lauren McCarran as substantive 51% Managing Member with operational role; Oceum itself as transferable infrastructure independent of any individual; NH-based engineering bench with partnered-contractor extension capacity for scope expansion. This is the operating model federal AI policy now contemplates for next-generation IT delivery, not a stage we are growing out of.

Engagement Models

Lakeshore Federal Services welcomes micro-purchase awards (under $10K), simplified acquisitions (under $250K), and sources sought / RFI responses to establish capability and past performance with new agency partners. We are particularly suited for AI/automation proofs-of-concept, modernization pilots, and discrete deliverable-based projects on Treasury and federal financial systems.