Agentic AI engineering
AI agents, orchestration frameworks, and coding harnesses built for consistency, reliability, and traceability. Persona engineering, dataset curation, and fine-tuned edge models.
TECHLEAD 187 LLC designs, builds, and securely deploys AI systems from edge to core for national security and commercial enterprises, with the evidence leadership needs to approve them.
$ tl187 mission --brief › map stakeholders, tech debt, outcomes MAPPED › build spec-driven agentic DevSecOps GREEN › gate prompt-injection scan, fail-closed ARMED › prove evals + evidence for ATO / cATO PASS › hand training + change management READY ✓ mission ready · human judgment in the loop
From concept through production: we build the agents, the platforms they run on, and the governance that lets regulated organizations trust them.
AI agents, orchestration frameworks, and coding harnesses built for consistency, reliability, and traceability. Persona engineering, dataset curation, and fine-tuned edge models.
Prompt-injection defense, fail-closed gating, and secure model access, delivered with the measurable evidence that carries a system through ATO and cATO.
Hybrid and multi-cloud built to reach operators in denied, degraded, and disconnected environments: Kubernetes, infrastructure as code, satellite, and 5G edge.
DataOps and MLOps platforms for classified enclaves, synthetic datasets, and knowledge-graph data platforms on edge hardware for predictive forecasting.
Hands-on workshops for every department, developers to back office. Responsible AI, secure workflows, and spec-driven development that sticks after we leave.
Proposals, statements of work, technical designs, and capability roadmaps. Principal-investigator leadership for R&D, with every recommendation mapped to ROI.
Explainable decisions, disciplined milestones, human judgment kept in the loop, and clean handoffs. Rigorous without becoming bureaucratic; security-first without slowing the mission.
Complexity, tech debt, tribal knowledge, stakeholders, and the business outcome before any technical direction.
Agentic DevSecOps and spec-driven development inside your existing infrastructure decisions.
Evaluations, security gates, and accreditation evidence that make every AI decision traceable.
Training, governance controls, and a change-management plan so your team owns the capability.
“Take the technology that wins next and make it usable by the force that fights now.”Operating principle · TECHLEAD 187
Live products and platforms built by TECHLEAD 187, in production today.
Far beyond a coding agent: a platform for powering applications with secure AI. Build agents and workflows, ground them on your own knowledge, and run them wherever your data is allowed to live, behind a fail-closed prompt-injection gate on every tool call. On local GPUs nothing leaves the host; with a cloud model, prompts and the code they include go to that provider, over network egress LUCID mediates and audits.
Ollama, llama.cpp, vLLM, or any OpenAI-compatible endpoint. Open-weight models on your hardware, fully air-gapped.
Anthropic, OpenAI, Google (Gemini 4 Argon), xAI (Grok 4.7), Amazon Bedrock, Google Vertex AI, and more. 47 current models, per-model cost showback, no lock-in.
Gov-lockdown routes every turn through the AskSage accredited gateway (FedRAMP High, IL5, IL6, Top Secret), with a CUI-isolated encrypted store and FIPS mode. Provider keys live in the OS vault, at-rest crypto runs through the OS FIPS 140 module, and a NIST SP 800-171 Rev 3 evidence package (control matrix, assessor guide, POA&M) ships with it.
Free, self-paced certification training: 11 curricula and 202 modules across generative AI, cloud architecture, cloud security, and CMMC CUI awareness, with printable contact-hour certificates.
Visit tacticalgenaitrainer.comA trading desk that AI agents can run: paper trading by default, a server-side risk gate, journal and lessons, Strategy Lab, six floors, 3D charts, and goals that write to your LUCID knowledge graph. New for insiders and contributors (Oct 26, 2026, in source ahead of the next installer release): Strategy Studio, plain-English rules turned into honest backtests with overfitting receipts and a robustness heatmap; the Market Map, a heatmap and screener that explains every result; and Smart Alerts that say why they fired, with a cost-aware history preview. All read-only: none of them can place an order. REST and 57 MCP tools for any agent; 365 offline tests.
Watch the 7:00 walkthroughMeeting capture without a bot in the room: local models, a signed journal, and no server holding your conversations.
Visit meetinghub.lucid-agents.comTwo planning engines from Black Arrow Systems, built by veterans for the military in partnership with TECHLEAD 187 LLC and the C.O.V.E.R.T. Consortium. The MDMP Engine puts the Army's Military Decision-Making Process in a staff officer's hands today; the Joint Planning Engine carries the same idea up to the joint force.
From mission receipt to OPORD, without the late-night doctrinal guesswork.
Live now · 3-day free trialA military decision-making process trainer that turns the seven-step MDMP into a clear, guided digital workflow. Real-time doctrinal coaching, automated IPB, interactive MIL-STD-2525D mapping, and one-click orders production in a single, secure web application. Built for staff rotations, exercises, ILE, and CGSC.
Legacy planning lives in static documents. Modern planning needs dynamic synchronization.
Seeking a government sponsorJPE is a secure, collaborative C2 application that guides military staffs through the Joint Planning Process. It offloads cognitive load and automates the heavy lifting of JIPOE, COA development, and wargaming, so planning cells can focus on what matters: out-thinking the adversary.
MDMP Engine and Joint Planning Engine are products of Black Arrow Systems LLC (owner Derek Hess), developed in partnership with TECHLEAD 187 LLC.
Not official U.S. Army or Department of War applications. Pricing as listed on tacticalmdmp.com.
Nick Chadwick hosts AI workshops through the C.O.V.E.R.T. Consortium, the veteran builder community he co-founded. Participants learn by building working applications with GenAI coding tools, backed by mentorship, demos, testing, and launch support.
A rank- and service-agnostic peer consortium for veterans, active duty, reserve, and guard members who deploy practical GenAI by doing, not talking, and transition into real technology careers. Apps built in the workshops, live today:
A full trainer for quantum communications, quantum computing and post-quantum security: 38 narrated 3D modules in six tracks, from a free math and physics primer to capstone missions. Real simulators run in the browser, from one qubit on the Bloch sphere and BB84 photon by photon over fiber with an eavesdropper on the line, to Shor's algorithm on up to twelve simulated qubits, ML-KEM and a CNSA 2.0 migration plan. Every constant cites its primary source. Each module plays like a film you can scrub and speed up, then practice with exact answers and a hint ladder, printable track certificates, and instructor reports for a whole class.
three.js · ElevenLabs narration · LUCID Agent plugin Open the app3:04 walkthroughA private learning companion for homeschooled kids who learn on Miacademy. A Chrome extension rides along with each lesson with hint ladders that guide without giving the answer away, and a small bridge on the family computer runs the Learning Hub: daily quests, a Mastery Path with a spaced Fix-it Bank, a Ready to learn next list, a Curriculum Map built on Marble's open skill taxonomy, two original 24-chapter story series narrated with a voice for every character and highlighted sentence by sentence, thirty-eight interactive 3D explainers with boxes kids can move or resize, and a Studio where kids prompt an agent to build their own. Records stay on the family computer.
LUCID Agent · ElevenLabs and local neural TTS See it on aiworkshopapps.com4:14 walkthroughEvidence-first career and relocation planner for the whole household: resume parsing, jobs near the target region for you and your partner, company reputation links and a one-click resume tailored to each opening (fact-checked, saved as an ATS-readable PDF), OSRM commute baselines, FEMA flood zones for choosing a home, verified housing and life-logistics research, volunteer and mentoring roles near the new home for each adult and the kids, three career paths on one budget with an AI ranking that shows its sources, and a decision brief red-teamed against its own assumptions.
LUCID Agent · NVIDIA DGX Spark Open the app5:26 walkthroughOne app for a household's health, insurance, and property tax on a local knowledge graph that stays on your computer. Voice-guided Care Finder searches plan doctor directories; Home Tax scores each taxing authority on a bill. New: health rules for every state with Florida county programs, your price rights and a playbook to negotiate a bill, life insurance with veteran options, PIN-encrypted settings, and data updates on request.
LUCID Agent · local voice AI Open the app3:37 walkthrough
Facility survey for security integrators, facilities and AEC teams, insurers, and public safety: on-site AI behind a CUI lockdown switch, native DWG drawings, IFC for Revit, and a one-click Autodesk package.
DGX Spark · DWG · Revit IFCNicholas Chadwick is a security-first technology executive whose 26-plus-year career spans enterprise network engineering, special operations cloud architecture, artificial intelligence fielded in combat, and first-of-their-kind government accreditations.
In 2026 he founded TECHLEAD 187 LLC to bring security-first agentic AI to national security and commercial enterprises. He also serves as a Senior Consultant with Infrastellis Advisory Group, solving compliance and automation challenges for government contractors and mapping every technical recommendation to a return-on-investment case leadership can act on.
As GenAI and Cloud Technical Lead at NT Concepts, he accredited a U.S. Air Force customer's first Google Cloud enclave at Impact Level 5, served as Principal Investigator for an Army advanced research program building AI agents on Google Gemini and Gemma for predictive logistics, and advised Army Futures Command on Next Generation Command and Control. Before that, as Senior Cloud Architect supporting U.S. Special Operations Command headquarters, he designed cloud-to-edge architectures for denied environments and, working directly with Google, deployed a containerized computer vision capability on remote edge systems.
In uniform, he is the senior Chief Warrant Officer of the AI Detachment in the Army Reserve's 75th Innovation Command, the Army's change agents for emerging technology, and was selected in 2024 and 2025 as an Army expert evaluator for next-generation Zero Trust, satellite communications, and AI-driven network resilience.
“Technology earns its place by changing outcomes people can measure.”
Eryl Chadwick leads go-to-market, sales, partnerships, and advertising for TECHLEAD 187 LLC, turning security-first agentic AI into offers that national security and commercial enterprises can evaluate and buy.
She has successfully led startups in the United States and the Philippines, and brings that cross-market operating experience to a company founded in 2026.
At TECHLEAD 187 she owns the go-to-market plan for LUCID Agent IDE, the Tactical GenAI Trainer, and the consulting practice: positioning, channel and technology partnerships, and the advertising that reaches the people who need secure AI. For sales, partnership, and reseller inquiries, email eryl@techlead187.com.
Twenty-six years delivering business and mission outcomes: risk measured before it is taken, mitigations engineered in from the start, and advanced technology transitioned rapidly from prototype to the people who depend on it.
Selected to evaluate next-generation Zero Trust, SATCOM, and AI/ML-driven network resilience (AutoPACE) solutions for the Cyber Capabilities Development Integration Directorate.
Joined the Army's change agents for emerging technology: S&T Advisory Board for Project Convergence, a Hacking4Defense project, a Tampa bug bounty, and synthetic VR/AR training in One World Terrain. Named AFCEA 40 Under 40 and AUSA Officer of the Year.
As communications lead, installed and ran the Army Research Laboratory's first AI-powered natural language translation capability, converting Arabic broadcast television to English in under a second, and networked it across a 350+ soldier task force during the Surge.
Learned networks as the soldier responsible for making them work in the field, then funded his education while serving, graduating with honors from the University of Toledo.
TECHLEAD 187 LLC is registered in SAM.gov and is a member of SOF RACER, the Special Operations Forces Rapid Acquisition Consortium for Emerging Requirements: a prototype Other Transaction consortium under 10 U.S.C. § 4022, managed by GovSignals, that moves SOF-peculiar capabilities from prototype to production.
Nick Chadwick's experience on prime contractor teams before founding TECHLEAD 187. These programs were performed by the prime contractors, not by TECHLEAD 187 LLC.
AI agents on Gemini and Gemma for Army predictive logistics, with knowledge graphs on edge hardware.
Principal Investigator · via NT ConceptsTechnical advisory for NGC2 Project Trinity with Army corps units.
Technical advisor and leadDefense Innovation Unit effort with AETC to modernize Air Force pilot training.
Tech lead and senior cloud architectEnterprise cloud and DevSecOps adoption across the SOF IT enterprise.
Senior cloud architectEnterprise-to-edge tactical support for Army Civil Affairs and Psychological Operations forces.
Systems engineerTECHLEAD 187 LLC is a Tampa, Florida engineering practice that designs, builds, and securely deploys agentic AI systems from edge to core for national security and commercial enterprises. Services cover agentic AI engineering, secure AI platforms and accreditation (RMF, IL5, FedRAMP, CMMC, Zero Trust), cloud-to-edge architecture, data and MLOps platforms, AI adoption workshops, and GovCon strategy.
LUCID Agent IDE (v2.5.2, released November 7, 2026) is TECHLEAD 187's secure-AI application platform. It builds agents and workflows behind a fail-closed prompt-injection gate and runs them on local GPUs (Ollama, llama.cpp, vLLM), in any enterprise cloud, or in CUI and Top Secret-authorized clouds through the AskSage accredited gateway. Download it at lucid-agents.com.
LUCID runs through them. With gov-lockdown on, every turn routes through the AskSage accredited gateway (FedRAMP High, IL5, IL6, Top Secret), with a CUI-isolated encrypted store and FIPS mode. The authorizations belong to the gateway and the accredited clouds, not to LUCID itself.
Prompt-injection defense, fail-closed gating, and secure model access, delivered with the measurable evidence that carries a system through ATO and cATO. Risk is measured before it is taken and mitigations are engineered in from the start.
Nicholas A. Chadwick, CISSP, is the founder of TECHLEAD 187 LLC. He has 26-plus years of U.S. Army and Army Reserve service, accredited a U.S. Air Force customer's first Google Cloud enclave at Impact Level 5, and in 2006 installed and ran the Army Research Laboratory's first AI-powered translation capability across a 350-plus soldier task force in Iraq.
Yes. Nick Chadwick hosts hands-on AI workshops through the C.O.V.E.R.T. Consortium, the veteran builder community he co-founded, where participants build working applications with GenAI coding tools. TECHLEAD 187 also publishes the Tactical GenAI Trainer and runs enterprise workshops for every department, developers to back office.
Tampa, Florida, serving national security and commercial enterprises nationwide.
Email nick@techlead187.com or connect with Nick Chadwick on LinkedIn. Tell us the mission and the outcome you need.
Agentic AI, accreditation, cloud-to-edge, or a team that needs to adopt AI safely. Tell us the mission and the outcome you need.
TAMPA, FLORIDA · SERVING NATIONAL SECURITY & COMMERCIAL ENTERPRISES NATIONWIDE