Federal Solutions
Zorost Intelligence is a small, disadvantaged business delivering production AI, data engineering, cloud modernization, and agentic systems to U.S. federal civilian, defense, and intelligence-community customers. SAM.gov registered. SDB-certified. 8(a) application in progress.
SAM.gov RegisteredSDB CertifiedPhD-Led14 CertificationsAir-Gap Capable
Engagement Vehicles & NAICS Codes
We engage federal customers through GSA MAS (under partner schedules), SBIR/STTR, subcontracting, OTAs and CSOs, set-asides (SDB; pursuing 8(a)), and direct award at the SAT level. We actively pursue teaming and mentor-protégé relationships with established primes.
Primary NAICS: 541715 (R&D in Physical, Engineering & Life Sciences) · 541511 (Custom Programming) · 541512 (Computer Systems Design) · 541513 (Facilities Management) · 541519 (Other Computer Services) · 518210 (Data Processing, Hosting) · 541611 / 541618 / 541690 (Management & Technical Consulting) · 541330 (Engineering Services) · 611420 (Computer Training).
Common PSC codes: D302, D307, D308, D310, R425, AC11–AC34.
Core Competencies
Applied AI & Machine Learning
Production AI systems combining causal inference, graph neural networks, conformal prediction, and multi-agent architectures. Fielded across regulated domains where calibration, traceability, and explainability are mandatory.
Generative & Agentic AI
Enterprise-grade LLM applications, multi-agent workflows, hybrid retrieval (vector + lexical), and document intelligence. Capable of cloud, hybrid, and air-gapped deployment with locally hosted models for environments without internet egress.
Cloud Modernization & Data Engineering
Azure, AWS, Databricks Lakehouse, and Microsoft Fabric. Migration of legacy reporting and analytics platforms to modern cloud-native architectures. Real-time streaming, batch, and unified governance via Unity Catalog.
Databricks Modernization Practice
Dedicated capability for legacy migration to Databricks: semantic layer reconstruction, ETL conversion, dimensional modeling, ML/AI on Mosaic AI, and governed reporting. See the standalone Databricks Modernization page →
Decision Support, Simulation & Digital Twins
Agent-based modeling, Monte Carlo simulation, scenario planning, and graph-based cascade modeling for operations and infrastructure resilience.
AI Governance & Responsible AI
NIST AI RMF–aligned governance, model risk management, bias and fairness testing, explainability, and audit-ready documentation — built and maintained on our own production systems before being recommended to customers.
Federal R&D Focus Areas
We pursue R&D under federal SBIR/STTR programs and CSO/OTA mechanisms. Public-facing focus areas (typical TRL 5–8 depending on topic):
- Causal AI for Operational Risk — structural causal models with sensitivity analysis for regulated decision support.
- Calibrated Forecasting at Scale — conformal prediction (LACP) and ensemble-meta-learner architectures for high-stakes scheduling and resource allocation.
- Retrieval-Augmented Knowledge Engines for Safety & Compliance Corpora — hybrid retrieval over very large public-domain safety document collections with citation-grounded generation.
- Air-Gapped Generative AI — local LLM stacks for environments without internet egress, with reproducible evaluation harnesses.
- Multi-Agent Decision Support — agentic architectures with tool-use, planner/executor separation, and human-in-the-loop checkpoints.
- Geospatial Intelligence Synthesis — causal graphs over geopolitical events with multi-source OSINT fusion.
Certifications
Federal & Program Management: Federal Contracting Officer’s Representative (COR) Level 1 · Project Management Professional (PMP) — PMI.
Cloud: Microsoft Azure Solutions Architect Expert · Azure Administrator Associate · Azure AI Engineer Associate · Fabric Analytics Engineer Associate · Fabric Data Engineer Associate · AWS AI Practitioner.
Databricks (five active): Data Engineer Associate · Data Engineer Professional · Machine Learning Professional · Generative AI Engineer Associate · Associate Developer for Apache Spark.
Education: PhD, Systems Engineering (Artificial Intelligence), George Washington University · Data Science, Harvard University (executive education) · Adjunct Professor, GWU (2020–Present).
Why Zorost Intelligence
We build what we advise. Our principal has personally engineered and deployed eleven production AI platforms across aviation, manufacturing, pharmaceutical research, compliance, finance, intelligence, and freight. Most agency AI partners produce strategy. We produce running systems with measurable calibration, monitored uptime, and audit trails.
Air-gapped capable. Several of our platforms run fully offline with local LLMs, local vector databases, and local model registries — appropriate for classified or otherwise disconnected environments.
Calibration-first practice. Our published methodology emphasizes calibrated probabilities (ECE) and conformal prediction intervals over headline accuracy — the standard a procurement officer should require for high-stakes decisions.
Small business agility, large-firm depth. SDB-certified small business. PhD-led. Fourteen active cloud and AI certifications.
Federal Contracting Inquiries
For contracting, teaming, and capability briefings:
Dr. Mehdi Hashemipour, PhD
Principal / Chief Executive — Zorost Intelligence LLC
Email: gov@zorost.com
Location: Washington, DC Metropolitan Area
SAM.gov Registered · SDB Certified · 8(a) Application in Progress
Specific customer references, contract numbers, and program-sensitive details are withheld from the public site by policy and provided to qualified contracting officers only.
