S&F AI Automation started as an internal R&D team, using AI to solve our own company's business problems. After building real expertise and a portfolio of delivered projects, we're now turning this into a practice that helps other businesses level-up their operations with AI.
We are a team of AI geeks who see our work as a way to express creativity. We are looking for someone who shares this obsession and will raise the bar: challenge our architecture decisions, bring tools and techniques we haven't tried, and help us turn prospects into projects we will enjoy engineering.
You'll join a small team at the stage where everything you do has outsized impact. This role brings senior-level engineering capacity to the team while also adding the presales muscle needed to grow the client base.
Technical requirements
- TypeScript (Node.js) OR Python — production-level proficiency in at least one.
- Building LLM-powered applications: agents, assistants, multi-step workflows, structured outputs. Prompt engineering: function calling / tool use, chain-of-thought. RAG: document ingestion, chunking, embedding models, relevance tuning. LLM SDKs: Anthropic Claude, OpenAI. Frameworks: LangChain, LangGraph. Building and integrating MCP (Model Context Protocol) servers.
- Agentic architecture: agent loops, tool orchestration, memory management, multi-agent systems. Knowing when agents are justified vs simpler alternatives.
- Workflow orchestration: Temporal (readiness to build durable workflows), n8n (low-code automation). Vector search: Qdrant, AWS OpenSearch, embedding pipelines, hybrid search.
- API & backend: FastAPI OR Fastify / NestJS. REST APIs, webhook integrations.
- AWS: Bedrock, SageMaker, Lambda, OpenSearch. Cloud ecosystems: AWS (Anthropic), Azure (OpenAI), GCP (Gemini) — platform-exclusive models and services. Docker, IaC basics, VPS provisioning, serverless patterns.
- Zero Data Retention policies, data isolation, private cloud deployments. LLM safety: guardrails, hallucination mitigation, prompt injection protection. Ability to articulate security posture to enterprise clients.
- LLM cost optimization: token budgeting, semantic caching, model routing, fallback strategies, choosing the right model tier.
- Active use of AI coding tools: Cursor, Claude Code, GitHub Copilot, or similar.
- Nice-to-have: Voice AI (STT/TTS), multimodal AI (OCR, vision-language models), LLM evaluation frameworks.
Who you are
You have 3+ years of experience in a technical presales, solutions engineering, or solutions architect role. You have led discovery calls, scoped proofs of concept, and handled technical objections in B2B sales cycles.
You have hands-on experience building AI-powered applications in a professional setting, not just personal experimentation.
Your English is fluent at C1+ level. You can lead live calls with English-speaking stakeholders and handle nuanced technical questions without hesitation.
You communicate consultatively. You guide stakeholders to clarity instead of overwhelming them with options. You can easily explain a complex concept to a non-technical audience.
You take ownership and initiative. You do not need a heavy playbook. You identify gaps, flag risks early, and move things forward. Much of the presale process is still being built, and you are comfortable shaping it as you go.
You are adaptable. You are comfortable in a small team where priorities shift, and where the role will evolve as the practice grows.
What the work looks like
Provide technical depth during presales.
You join prospect calls alongside the department head and own the technical side of the conversation.You probe for technical constraints, assess feasibility in real time, identify integration points or data dependencies, and start forming an architecture before the call ends.
Design AI solutions that match the actual problem.
You translate business context into an architecture with realistic scope and effort estimates. You know when prompt engineering is enough, when RAG is needed, when fine tuning or custom development is justified, and you can explain why.
Connect AI to the tools businesses already use.
You design integrations that make AI solutions practical: pulling data from CRMs or ERPs, syncing with cloud drives or email, pushing outputs into Slack channels or project management tools. You think in terms of end to end workflows, not standalone modules.
Contribute to building, not just sales.
Between engagements, you write production code, guide architecture decisions, and help the team ship client and internal projects. This role carries engineering weight, not just advisory influence.
Outcomes over the next 6–12 months
Strengthen the team’s engineering capability
Provide architectural guidance on client or internal projects so solutions are more reliable and scalable than what would have been built without your input.
Ramp up as a presales partner within 30 days
Independently lead the technical portion of prospect calls by the end of month one. The department head should be confident you ask strong discovery questions, communicate clearly, and propose relevant approaches without heavy coaching or supervision.
Help close the first $80k in new revenue
Participate in presales engagements that result in at least $80k in signed project value within the first six months. This includes Discovery Sprints and follow-on projects that come from them.
About Speed & Function
Speed & Function is a software development company focused on long-term partnerships, not transactional outsourcing. Since 2006, we’ve been building digital products and teams with a strong emphasis on trust, ownership, and collaboration. We evolve our processes as real-world complexity demands.
How to apply
Apply with a short cover letter that includes why you’re considering a new opportunity, your hourly rate, your English level, your availability, and your location.