About Command|Link
Command|Link is a global SaaS Platform providing network, voice services, and IT security solutions, helping corporations consolidate their core infrastructure into a single vendor and layering on a proprietary single pane of glass platform. Command|Link has revolutionized the IT industry by tackling the problems our competitors create. In recognition for our unprecedented innovation and dedication, Command|Link was recognized as the SD-WAN Product of the Year, ITSM Visionary Spotlight, UCaaS Product of the Year, NaaS Product of the Year, Supplier of the Year, and the AT&T Strategic Growth Partner. Command|Link has built the only IT platform for scale that solves ISP vendor sprawl and IT headaches. We make it easy for our customers to get more done, maximize uptime and improve the bottom line.
Learn more about us here!
This is a 100% remote position
About your new role:
Discovery and automated topology are the single biggest unmet need in monitoring today, and a core piece of CommandLink's platform differentiation. This role brings applied data science, the engineering discipline to run it in production, and a working grounding in networking and telemetry protocols to that problem and to the broader set of data challenges the Classification and Discovery team encounters as it turns raw telemetry into a living, queryable map of customer infrastructure.
As a Senior Data Scientist, you will build the models and evaluation frameworks that resolve entities, score confidence in candidate matches, and separate reliable signal from noise, and you will own those models end to end once they ship. That work depends on understanding what the underlying data actually represents: you need to reason about what a syslog message, an SNMP trap, or an OpenTelemetry span tells you about a real system, not just treat telemetry as an abstract feature set. Your work extends past the classification layer itself. You will partner with adjacent teams in security, alerting, and observability to apply statistical thinking and modeling to problems across the platform, from anomaly detection to alert quality to telemetry pattern recognition.
This is a hands-on, production-first role. You will spend as much time in the codebase, in streaming pipelines, and monitoring model health in production as you spend in exploratory analysis and model development.
Key Responsibilities:
-
Design, build, and ship models for entity resolution and deduplication directly into production pipelines, including confidence scoring approaches that distinguish reliable matches from ambiguous ones.
-
Own the full lifecycle of models you build in production: deployment, monitoring, retraining triggers, and incident response when model performance degrades.
-
Build and maintain the model-serving and evaluation infrastructure itself, including precision and recall tracking, confidence calibration, and drift detection running continuously against live traffic.
-
Write production-grade code, not just notebooks, and work directly within streaming and graph-based systems such as Kafka and Memgraph to get models into the platform's real-time pipelines.
-
Reason directly about network and telemetry protocol data, including syslog, SNMP, and OpenTelemetry, to understand what raw signal actually represents before it becomes a model input.
-
Partner with the Classification and Discovery team to bring statistical rigor and applied machine learning to how the platform identifies, classifies, and resolves entities across customer infrastructure.
-
Extend data science methods beyond the classification layer, partnering with security, alerting, and observability teams to apply modeling and analytics to broader platform problems such as anomaly detection, alert quality, and telemetry pattern recognition.
-
Analyze large volumes of telemetry, log, and topology data to identify patterns that inform product and engineering decisions.
-
Apply sound software engineering practices to model code, including version control, testing, and participation in code review, so models are maintainable by the broader engineering team.
-
Communicate findings and recommendations clearly to both technical and non-technical stakeholders, shaping how the team prioritizes data-driven improvements.
-
Mentor other data scientists and analysts on best practices for model development, evaluation, and running models reliably in production.
-
Takes on additional responsibilities and projects as needed to support the success of the team and organization.
What you'll need for success:
Required:
-
Strong applied statistics and machine learning background, with hands-on experience deploying and operating models in production, not just building them offline.
-
Strong production Python skills, including experience writing tested, maintainable code and working directly in a shared codebase alongside engineers.
-
Direct, hands-on experience with Kafka and streaming or event-driven systems, sufficient to build and debug pipelines yourself rather than hand work off.
-
Direct, hands-on experience with a graph database such as Memgraph or Neo4j, including writing and optimizing queries against production data.
-
Working knowledge of networking fundamentals and common telemetry protocols such as SNMP, syslog, and OpenTelemetry, along with experience working with infrastructure or observability data.
-
Experience with entity resolution, record linkage, or deduplication techniques, such as Splink or a comparable framework, applied to real-world messy data.
-
Experience building monitoring and evaluation systems for models already in production, including precision, recall, confidence calibration, and drift detection.
-
Comfort working with large-scale telemetry, log, or event data, and translating noisy real-world signals into structured, reliable model inputs.
-
Strong communication skills, with the ability to translate data science findings into concrete product and engineering decisions.
-
Track record of owning models through their full production lifecycle, from initial deployment through ongoing operation and iteration.
Nice to Have:
-
Experience with MLOps tooling for model versioning, deployment automation, or CI/CD pipelines for machine learning.
-
Familiarity with providing structured context to LLMs for reasoning over topology, troubleshooting, or remediation workflows.
-
Experience with anomaly detection or forecasting applied to operational or monitoring data.
-
Background in cybersecurity, network detection and response, or infrastructure observability products.
Why you'll love life at Command|Link
Join us at CommandLink, where you'll have the opportunity to shape the future of business communication. We value the innovative spirit and seek individuals ready to bring their unique vision and expertise to a team that values bold ideas and strategic thinking. Are you ready to make an impact?
-
Room to grow at a high-growth company
-
An environment that celebrates ideas and innovation
-
Your work will have a tangible impact
-
Flexible time off
-
Fun events at cool locations
-
Employee referral bonuses to encourage the addition of great new people to the team
At CommandLink, we’re committed to creating a fair, consistent, and efficient hiring experience. As part of our process, we use AI-assisted tools to help review and analyze applications. These tools support our recruiting team by identifying qualifications and experience that align with the requirements of each role.
AI tools are used only to assist in the evaluation process — they do not make final hiring decisions. Every application is reviewed by a member of our recruiting or hiring team before any decisions are made.
