Jelena Ciaralli

Strategic operator. Builds the systems that make AI work in the day-to-day.

jelena.ciaralli@gmail.com | linkedin.com/in/jelena-ciaralli | Jacksonville, FL | Remote US
Strategic & Business Operations Cross-Functional Initiatives AI Tools in Daily Workflows 0-to-1 Process Design Stakeholder Management High-Growth Startup Operations
View Resume
What I do

I build the operational backbone that lets small teams ship like big ones. Cross-functional initiatives. 0-to-1 process design. AI tools embedded into daily workflows so the team scales without adding headcount.

13+ years across SaaS, fintech, and proptech. Most recent chapter at Rayse (early-stage proptech): first strategic operations hire, built ops from zero, embedded LLM workflows into reporting, partner comms, and project tracking.

Featured Work

Applied AI prototype — Call-center post-call assistant

Built April 2026 for a Series B fintech client. Synthetic dataset. 6-hour build.

Problem. Call-center agents spend the minutes after every call on case notes, disposition coding, and routing. Repetitive, error-prone, gets compressed when call volume spikes.

What I built. A functional prototype that takes a call transcript and returns three things in one pass:

  • Structured case note. Summary, key facts, customer intent.
  • Disposition code with confidence score.
  • Recommended routing or follow-up action.

Inputs. Synthetic dataset of 300 calls across 10 fields. Generated end-to-end. Verified for shape before feeding through.

Design choices.

  • Input schema locked before the prompt. Output rubric defined before testing. Forces the model to fit the workflow, not the other way around.
  • Mandatory human-in-the-loop review surface. The output is a draft, not the final record.
  • Tool transparency baked into the brief. Documented exactly which AI tools touched which step.
What it demonstrates. How I think about applied AI. Schema first. Drift prevention day one. Builder + operator hybrid.

Loom walkthrough covers the design choices, the prototype in action, and the failure modes I planned for. Loom link available on request.

AI-augmented operating system at Rayse

Aug 2024 – Apr 2026. First strategic operations hire at a Series A proptech.

Problem. Lean team. Growing partner pipeline. No SOPs, no playbooks, no operating cadence. Founders running on tribal knowledge and ad hoc execution.

What I built.

  • Operating cadence. Mon/Wed/Fri rhythm with clear inputs and outputs per session.
  • 3-tier escalation framework so issues stopped landing on the CEO's desk by default.
  • Project archetyping so new work could be scoped against repeatable templates.
  • Notion portfolio with multi-views by partner, by stage, by owner.
  • Retro process that fed back into the templates each cycle.

Where AI came in (last few months of tenure).

  • LLM workflows for partner reporting. Input schema, output rubric, human review before send.
  • Stakeholder communication drafting with style controls so the voice stayed consistent.
  • Project tracking and meeting synthesis so the team spent more time on partners and less on admin.
Results. 12 partner launches executed end-to-end. ~$50K/year in vendor cost cut through a usage-data-first tooling audit. Enabled the team to operate at significantly higher capacity without adding ops headcount. Trained non-AI-fluent teammates so the workflows survived past the original builder.

Proprietary CRM build at Enspire for Enterprise

Mar 2020 – Mar 2022. Product Owner / Product Manager.

Problem. Paying ~$60K/year for a third-party CRM that didn't fit the business. Reporting disconnected from the ad and call-tracking data the team actually used.

What I led.

  • Roadmap and backlog for a proprietary CRM across three agile engineering teams.
  • Reporting dashboards that turned ad + call-tracking data into actionable insight for senior leadership.
  • Internal tooling that connected production, onboarding, and customer experience.
Results. ~$60K/year saved by replacing third-party tooling. ~$10K/month in new recurring revenue from integrated reporting. Production time down 20%. Customer onboarding time down 35%. Dashboard engagement up 25%.
How I work
Schema first.
I lock the inputs and outputs before the prompt or the process. If the shape isn't clear, the workflow drifts.
Human-in-the-loop by default.
AI drafts, humans ship. Especially when the audience is a partner, a customer, or a senior leader.
Build for the next operator.
I write SOPs and enablement assets as if I'm handing them to someone less experienced. That's the real test.
Operating cadence beats heroics.
A weekly rhythm with clear inputs catches more than ad hoc fire drills.
Audits start with usage data, not opinions.
Vendor and tooling decisions follow what people actually use, not what they remember using.
Skills + Tools
Strategic & Business Operations. Cross-functional initiatives, process design, project scoping, stakeholder management, vendor and contract strategy, operating cadence design.
AI + Automation. LLM workflow design, prompt + schema engineering, human-in-the-loop quality review, AI enablement and training, Claude, ChatGPT.
Customer Experience + GTM. Customer onboarding, technical onboarding (SSO, API, data integrations), partner launches, adoption + engagement.
Systems. Jira, Confluence, Notion, Monday.com, Slack, Google Workspace. Smartsheet (Web.com 2019-2020). Proprietary CRM (built at Enspire).
Notes for hiring managers

Targeting Strategic Operations, Business Operations, and Applied AI roles. Remote US or hybrid in Jacksonville. Comp expectations align with senior IC and manager bands for those roles.

References available on request. Resume available here or via LinkedIn.