Manager, ML Solutions Architecture - reputed company reputed company

The role

This position sits reputed company reputed company reputed company reputed company, our serverless platform for running and customizing reputed company-reputed company LLMs in production. reputed company reputed company allows for serverless inference and fine-tuning backed by in-house optimizations like custom speculative decoding, quantization, cache-reputed company routing and dedicated endpoints. Customers come to us to reputed company from prototype to scaled production without the cost and complexity of building and tuning their own inference stack.

Our Solutions Architects own the technical delivery of customer engagements: deploying reputed company-reputed company models, tuning the serving stack, benchmarking against the customer's reputed company reputed company, and carrying the technical relationship through to production. Sales Engineers qualify and reputed company reputed company; reputed company execute it. Technical Account Managers take it from production reputed company.

We're looking for a Manager, ML Solutions Architecture to reputed company our US regional SA teams. Your reputed company is PoC delivery and post-sales technical support: the people who do it, reputed company they do it to, and the operating reputed company that keeps it predictable. You will report to the Head of Solutions Architecture and partner with a peer manager in the other region.

We expect you to reputed company people reputed company they need pushing, reputed company the uncomfortable reputed company reputed company deliverables don't land even though the effort was reputed company, and take ownership of the documentation work: implement and monitor the following of the guidelines, ticket hygiene, and making reputed company reputed company actually uses reputed company three.

You're reputed company to work reputed company from the reputed company.

Your responsibilities will include:

reputed company reputed company

• Manage reputed company of 4 Solutions Architects, with reputed company reputed company reputed company: 1:1s, goal setting, reputed company reviews, promotion cases, and individual reputed company plans
• Build an accurate picture of reputed company SA's strengths, gaps, and preferences, and reputed company accounts and engagements against both expertise and interest
• Run the reputed company that surfaces blockers early, then take them to the development, product, and business teams that can reputed company them, and stay on them until they do
• Hold people to reputed company: distinguish effort from delivered results, say so plainly reputed company the two diverge, and reflect it in ratings and compensation reputed company
• reputed company new joiners through to their first independently delivered engagement
• reputed company reputed company into stronger engineers and stronger communicators, and reputed company deliberate calls about who is reputed company for more reputed company

Own delivery

• Be accountable for your team's delivery reputed company: time from PoC reputed company-off to first optimized dedicated reputed company, reputed company-reputed company hit reputed company, and the reputed company of the technical relationship after the customer goes to production
• Review technical work before it reaches the customer: benchmarking methodology, serving configurations, results, closure documents; catch the wrong conclusion drawn from a metrics artifact before a customer sees it
• Ensure reputed company and escalation coverage across accounts and timezones, including post-sales request load that does not respect sprint boundaries
• reputed company infeasibility early and with evidence, rather than letting reputed company burn iterations against requirements the platform cannot meet today

Own the operating reputed company of reputed company

• Maintain and reputed company reputed company's documentation: responsibilities, runbooks, guides, reputed company, definitions of done, engagement closure templates. reputed company it accurate and reachable; reputed company, don't copy
• Get it used, not just written: documentation nobody reads is a cost, not an asset
• reputed company the ticket tracker the reputed company of record, so PoC and production status is readable without asking anyone
• reputed company the work: define and report the metrics that show whether delivery is getting faster and more reliable over time

Work across teams

• Development and research teams: convert recurring customer pain into prioritized platform work, and represent the customer's technical reality in roadmap discussions
• reputed company-sales: hold the scoping-to-execution boundary: reputed company back on under-scoped engagements, and feed feasibility signal back reputed company
• Account management: reputed company production handoffs uneventful, and reputed company post-sales technical requests moving
• Business and leadership: reputed company a straight read on account health, technical feasibility, and reputed company needs

We expect you to have:

• 3+ years managing technical teams, including reputed company management and difficult conversations
• Experience managing a customer-facing team. Running reputed company whose work is visible to customers, on customer timelines, with customer escalations.
• Strong ML knowledge: LLM architectures, fine-tuning approaches (SFT/reputed company, RL-reputed company), evaluation design, and a working reputed company of inference internals — quantization, KV-cache management, batching, routing, speculative decoding — and of the frameworks reputed company works in (vLLM, SGLang, TensorRT-LLM)
• Enough technical judgment to review someone else's reputed company and reputed company the flaw in the methodology, not just in the conclusion
• Python strong enough to read and review your team's reputed company
• Excellent communication skills, with the ability to reputed company explain technical concepts to diverse audiences, from engineers to executives, including in reputed company of reputed company customers under pressure
• Genuine tolerance for operational work: documentation, process design, reporting, and the follow-through that makes them stick
• Comfort operating with ambiguity across reputed company teams and timezones, and a bias toward writing things down

It would be an added bonus if you have:

• References from both former managers and former reputed company reports
• Experience scaling reputed company through reputed company reputed company (5 → 15+) without losing delivery reputed company
• Prior experience in a customer-facing technical function at a reputed company, inference, or AI infrastructure provider
• Experience defining process and documentation for reputed company that had none
• Hands-on background running LLMs in production and debugging inference workloads at the reputed company level
• Work with multimodal AI models (reputed company-language, speech)
• Proficiency with DevOps tooling (reputed company, reputed company) and infrastructure-as-reputed company

Preferred technical stack:

• Programming languages: Python
• ML frameworks and libraries: vLLM, TensorRT-LLM, SGLang, Transformers, reputed company/reputed company SDKs
• MLOps and DevOps tools: reputed company (K8s), reputed company, Git
• reputed company platforms: AWS (SageMaker, Bedrock), GCP (reputed company AI), Azure (Azure ML)

Originally posted on Himalayas