AI/ML Solutions Architect

Provectus·Colombia·Удалённо·1 мес. назад

As an ML Solutions Architect, you'll be the technical bridge between clients and delivery teams. You'll lead pre-sales technical discussions, design ML architectures that solve business problems, and ensure solutions are feasible, scalable, and aligned with client needs. This is a highly client-facing role requiring both deep technical expertise and strong communication skills.

In the era of Generative AI and autonomous systems, you'll also be responsible for architecting agentic solutions that leverage LLMs, tool ecosystems, and AI-assisted workflows to deliver transformative value to clients.

Core Responsibilities: 1. Pre-Sales and Solution Design (45%):

Lead technical discovery sessions with prospective clients
Understand client business problems and translate them into ML solutions
Design end-to-end ML architectures and technical proposals
Create compelling technical presentations and demonstrations
Estimate project scope, timelines, cost, and resource requirements
Support General Managers in winning new business

2. Client-Facing Technical Leadership (25%):

Serve as the primary technical point of contact for clients

Manage technical stakeholder expectations

Present technical solutions to both technical and non-technical audiences

Navigate complex organizational dynamics and conflicting priorities

Ensure client satisfaction throughout the project lifecycle

Build long-term trusted advisor relationship

3. Agentic Solutions Architecture (15%)

Architect agentic AI solutions that leverage autonomous decision-making and tool orchestration
Design MCP (Model Context Protocol) integration strategies for client environments
Evaluate and recommend appropriate agent frameworks (LangGraph, Claude Agent SDK, etc.) for client use cases
Create POC demonstrations showcasing agentic capabilities using AI-assisted development tools
Advise clients on build vs. buy decisions for agentic components
Develop reference architectures for common agentic patterns (RAG agents, multi-agent systems, tool-using agents)
Assess AgentOps requirements including monitoring, evaluation, and cost optimization

4. Internal Collaboration and Handoff (15%):

Collaborate with delivery teams to ensure smooth handoff

Provide technical guidance during project execution

Contribute to the development of reusable solution patterns and agentic accelerators

Share learnings and best practices with ML practice

Mentor engineers on client communication and solution design

Contribute to Provectus AI toolkit documentation and solution template

Requirements: 1. ML Architecture and Design

Solution Design: Ability to architect end-to-end ML systems for diverse business problems

ML Lifecycle: Deep understanding of the full ML lifecycle from data to deployment

System Design: Experience designing scalable, production-grade ML architectures

Trade-off Analysis: Ability to evaluate technical approaches (cost, performance, complexity)

Feasibility Assessment: Quickly assess if ML is an appropriate solution for a proble

2. Agentic Engineering & AI-Assisted Development:

Agentic Architecture: Deep understanding of agent design patterns, state management, and orchestration frameworks
Claude Ecosystem: Hands-on experience with Claude Code, Claude Agent SDK, and Anthropic's tool ecosystem
MCP Proficiency: Understanding of Model Context Protocol architecture for designing client integrations
Agent Frameworks: Practical knowledge of LangGraph, LangChain agents, and multi-agent orchestration patterns
AI-Assisted Workflows: Demonstrated experience with AI coding assistants (Cursor, GitHub Copilot, Claude Code) for rapid prototyping
Tool Ecosystem Design: Ability to architect function calling and tool use strategies for complex client requirements
AgentOps Understanding: Knowledge of agent monitoring, evaluation frameworks, and cost optimization strategies
POC Development: Ability to rapidly build compelling agentic demonstrations using AI-assisted development

3. ML Breadth

Multiple ML Domains: Experience across various ML applications (RAG, Computer Vision, Time Series, Recommendation, etc.)

LLM Solutions: Strong experience in architecting LLM-based applications including agentic systems

Classical ML: Foundation in traditional ML algorithms and when to use them

Deep Learning: Understanding of neural network architectures and applications

MLOps/LLMOps/AgentOps: Knowledge of production ML infrastructure and DevOps practices for all ML paradigms

4. Cloud and Infrastructure

AWS Expertise: Advanced knowledge of AWS ML and data services (SageMaker, Bedrock, Lambda, ECS, etc.)

Amazon Bedrock: Deep understanding of Bedrock agents, knowledge bases, and model hosting options

Multi-Cloud Awareness: Understanding of Azure, GCP alternatives for comparative discussions

Serverless Architectures: Experience with Lambda, API Gateway, Step Functions for agentic workflows

Cost Optimization: Ability to design cost-effective solutions with clear TCO analysis

Security and Compliance: Understanding of data security, privacy, and compliance requirements

5. Data Architecture

Data Pipelines: Understanding of ETL/ELT patterns and tools

Data Storage: Knowledge of databases, data lakes, vector databases, and warehouses

Data Quality: Understanding of data validation and monitoring

Real-time vs Batch: Ability to design for different data processing needs

Nice-to-Have Technical Skills

AWS Certifications (Solutions Architect Professional, ML Specialty)
Experience with specific industries (Finance, Healthcare, Retail, etc.)
Knowledge of AI ethics and responsible AI practices
Experience with edge ML and IoT deployments
Published thought leadership (blogs, talks, whitepapers)
Contributions to open-source agent frameworks or MCP servers

Experience and Education:

Demonstrated competency equivalent to 6-8+ years in ML/data science roles
Proven track record in client-facing technical roles
Experience leading pre-sales or discovery engagements
Portfolio of successfully architected and delivered ML solutions
History of winning business through technical leadership
Demonstrated experience with agentic AI architectures and AI-assisted development workflows
Education: Bachelor's or Master's degree in Computer Science, Data Science, Engineering, or related technical field or Equivalent experience with strong technical foundation and demonstrable expertise