Head of AI Initiatives

Files.com•Austin, TX

9d•Onsite

About The Position

At Files.com, our mission is to accelerate the world’s success by building and distributing technology that makes data integration and movement effortless. We build mission-critical infrastructure that powers complex, high-stakes digital supply chains for thousands of organizations. We’re creating a completely new role at Files.com: Head of AI Initiatives. This isn’t a research job. It isn’t an internal “AI strategy” role. And it definitely isn’t an innovation lab that ships demos. This is a builder–operator seat for someone who wants to embed AI into the operating system of a serious SaaS business — across sales, marketing, support, engineering, and operations — and ship real, margin-positive AI systems that people use every day. AI at Files.com: From Hype to Real Tools We’re building real, operational AI infrastructure that’s live today — powering both customer-facing features and internal workflows. Customer-Facing AI Initiatives MCP Integration (Live Now): We released our initial MCP (Model Context Protocol) integration last year. Customers can execute agentic tasks — automated sequences of actions — through their preferred AI stack. In-App Agentic AI Experience (Coming Soon): We’re building an in-app version of the above, where customers can trigger agentic workflows directly inside Files.com. In this version, we’ll manage the AI infrastructure behind the scenes, dramatically reducing time-to-value. AI-Powered ETL Infrastructure (In Development): We’re expanding customer-facing AI into AI-powered Extract/Transform/Load inside the Files.com UI: summarizing files, extracting data, transforming documents, and enabling enterprise customers to run highly customized workflows using their own data within Files.com. Internal AI Tools (All Live and In Use Today) We’re aggressively deploying AI inside our business to learn what works firsthand — and to build credibility, training data, and internal reference architectures that fuel the external roadmap. Already launched: Transcription + Contextual Wrapping Every Zoom and phone call across key teams is transcribed automatically — but unlike commodity transcription, we feed Files.com-specific context (product names, acronyms, terminology) to materially improve recognition accuracy. We own the ingestion pipeline end-to-end: upload, batching, retries, metadata tagging, and permissioning. Transcription Refinement Engine A second tool strips fluff (greetings, “ums,” connection issues) and shrinks transcripts ~80% with no meaningful loss — lowering AI costs and improving downstream accuracy. Sales Brief Generator A tool that digests every prospect conversation and generates a complete internal deal brief for weekly pipeline review — cutting AE prep time by 50%+ and helping scale sales without scaling headcount. Coaching & Gap Analysis Engine Custom evaluation prompts/models flag missed follow-ups, incomplete answers, tone issues, and process adherence gaps across onboarding, support, and sales transcripts. Managers jump directly to high-signal coaching moments. And unlike Gong-style black boxes, we control the prompts and inject our training philosophy into the system. Next in Development: AI-Powered Upsell Detection We’re building what may become our highest-impact internal AI tool: identify every expansion opportunity across customer communications (calls + email). The hard part isn’t “use GPT-4.” It’s building the specialized ETL pipeline to normalize, structure, and index messy comms data from Gmail, Zoom, CRM, and archives — then layering AI on top. This also becomes a reference architecture we can productize as part of an ELT + AI bundle. The Role and Its Impact Today we have roughly 10 production AI initiatives. Our goal is to build 50–100. “Production” means persistent, systematized, observable, and teachable — not ad hoc prompting. You will own the outcome: turning AI into repeatable leverage across the company, through a portfolio of high-ROI initiatives — and building the standards, libraries, infrastructure, and enablement needed for that to compound. This role is for someone who wants to be the internal AI founder/operator — shipping constantly, teaching constantly, and raising the company’s baseline capability permanently.

Requirements

You have strong, practical AI depth
You understand the current model landscape and how to evaluate tradeoffs
You’ve deployed AI in real environments where reliability, costs, and governance matter
You know what actually drives adoption inside companies
You’re a doer, not a delegator
You can build: prompts, workflows, tooling, lightweight platforms
You can move from ambiguity to shipping without needing perfect requirements
You enjoy iterating based on real feedback and metrics
You think in systems and compounding leverage
You love turning chaos into repeatable workflows
You build libraries, playbooks, and “default paths” that reduce variance
You care about durability, maintainability, and upgrade paths
You’re a teacher and multiplier
You can train and coach people with very different skill levels
You make advanced concepts practical and approachable
You leave behind capability, not dependence
You have high standards for production
You care about observability, security, data handling, and lifecycle management
You don’t confuse “cool” with “useful”
You know how to balance speed with correctness

Responsibilities

Build and Scale a Portfolio of AI Initiatives
Identify high-leverage opportunities across every function
Prioritize ruthlessly for ROI, speed-to-production, and durability
Deliver dozens of initiatives that save time, reduce cognitive load, and increase quality
Ship End-to-End Systems (Hands-On)
Prototype quickly, then productionize what works
Integrate AI into the tools we already use (CRM, support tooling, docs, repos, analytics, internal systems)
Instrument usage and outcomes; iterate relentlessly
Own the AI Stack and Model Strategy
Evaluate frontier vs cost-efficient models and tools
Make smart decisions about where we pay for premium capability
Avoid lock-in by designing swap-ready abstractions and clean interfaces
Build Prompt + Workflow Libraries (Company IP)
Create reusable templates, chains, and role-based playbooks
Version and maintain them like real software
Make the “best way” easy and repeatable
Build Knowledge/Context Systems
Implement RAG systems that respect permissions and source-of-truth hygiene
Improve internal documentation discipline by making knowledge usable
Ensure context is accurate, fresh, and governed
Teach and Enable the Whole Company
Run workshops, office hours, and hands-on training
Produce a high volume of practical internal content
Build internal champions so capability becomes distributed, not centralized
Create Governance Without Bureaucracy
Establish practical rules for safe/approved AI usage
Prevent tool sprawl, data leaks, and chaos — while enabling “fast yes”
Partner with engineering/security to ensure reliability and compliance