Networks built the way hyperscalers build them — for federal missions.
InfraSmith Technologies is a Northern Virginia small business led by a technical leader from Google, Meta, AWS, and Microsoft — bringing principal-level network engineering, AI/HPC cluster fabrics, zero trust, OT security, and AI-driven operations to federal programs.
Principal-level engineering, end to end.
Network architecture & O&M
Enterprise, campus, WAN, and data center design, engineering, and operations across Layer 2/3.
AI/HPC infrastructure networking
Engineering, deployment, and operations for AI/ML training and inference cluster fabrics, including NVIDIA GB200-class rack-scale GPU systems — automated repair raising first-time diagnosis accuracy from ~30% to 70%+ and cutting mean-time-to-repair by 70%+.
Automation & IaC
Network automation and infrastructure-as-code: configuration generation, network modeling, qualification testing, and auto-mitigation.
Zero trust & segmentation
Zero trust network implementation, segmentation and micro-segmentation policy design and rollout.
OT & critical infrastructure
Design and security of OT and industrial control networks for critical infrastructure and data center facilities.
Cloud provider networking
Network infrastructure for planet-scale cloud services — provider-side engineering and operations at AWS and Microsoft — plus SD-WAN and DNS/DHCP/IPAM engineering.
Led where the world's largest networks are built.
The founder has led platform programs from concept to production, directed and mentored engineering teams, and shaped hardware and software vendor roadmaps — at four hyperscalers whose networks serve billions of users daily.
Technical lead for qualification and global deployment of next-generation backbone routing platforms, including first-in-network 400G product introduction; leads engineers delivering secure, mission-critical facility network programs.
Led AI-network repair automation for GPU training clusters, including NVIDIA GB200-class rack-scale systems — deployment automation & MOPs, fabric-design collaboration, and auto-mitigation raising first-time diagnosis accuracy from ~30% to 70%+ and cutting mean-time-to-repair by 70%+ while protecting GPU capacity.
Designed data center network fabrics and next-generation rack architectures for planet-scale storage services; presented designs through review boards to production.
Data center and service networking design, implementation, and troubleshooting for global online services.
Founder: Muhammad Tahir Malik · Founder & Principal Engineer · CCIE R&S #34578 (inactive). InfraSmith was founded in 2026 in Northern Virginia to bring that discipline to the federal mission. Corporate experience described at capability level; specifics available in conversation.
Small-business speed. Hyperscaler command.
Principal-led delivery with senior-only engineering benches as engagements scale. Every project runs with the operational discipline of the world's largest networks: telemetry-driven monitoring, capacity planning, runbooks, configuration and change management, and documentation-complete handoffs. We exist to make primes and program offices look good on the network portions of their missions — and we mobilize fast from Northern Virginia.
Built to slot into your team.
For large primes
Every subcontract dollar to InfraSmith counts toward your small business subcontracting goals under FAR 52.219-8/-9 — while adding genuinely rare principal-level network capability to your delivery team.
For small primes
As a small business subcontractor, InfraSmith is a "similarly situated entity" — our work counts toward your required self-performance under the FAR limitations on subcontracting, letting you bid network scope you couldn't staff alone.
How we engage
Corp-to-corp subcontracting and teaming · T&M or Firm Fixed Price · immediate mobilization from Northern Virginia · W-9, Certificate of Insurance, and references available upon request.
Registered, verifiable, ready to team.
Add rare network capability to your team.
Seeking federal subcontracting and teaming opportunities on network infrastructure, AI/HPC cluster fabrics, modernization, zero trust, and OT security work. Capability statement available on request.