Worked on building core products, infrastructure, and internal tools
Role
Software Engineer
Timeline
May 2025 - May 2026
Team
1 Project Manager
8 Engineers
Tools
AWS
Next.js
PostgreSQL
Supabase
Overview
RedRover is a Reddit marketing platform powered by AI agents that automate finding relevant threads and posting content to drive traffic and leads. RedRover was built on the belief that distribution drives sales, and sales drive revenue. I joined Clover Labs to help make RedRover the best agentic marketing platform.
Strategy Reports
RedRover onboards clients across a lot of industries, so we needed a consistent way to understand a client's market and turn that into an actionable Reddit plan. I built the strategy report generator used during onboarding: it takes a client URL plus a small amount of context (industry and positioning) and produces a shareable report that outlines where the client should show up on Reddit and what they should say.
The report standardizes the core outputs: recommended subreddits (by intent), high-intent keywords, competitor set, a TOF/MOF/BOF content plan with example angles, and a clear summary of monthly deliverables. This replaced our manual onboarding process and saved our customer-success agents tons of time.
App Store Integration
To give clients visibility into app performance without leaving the product, I built an integration with the App Store Connect API that let customers connect their Apple developer account and pull down key metrics. The goal was a low-friction setup flow that worked across different client account configurations, while keeping the integration reliable and safe (auth, permissions, and failure handling).
On the product side, I turned the raw App Store Connect data into download trend views and metrics dashboards that were actually usable: time-series charts, comparisons over time, and clean summaries that answered "how are we doing?" at a glance. End-to-end, this meant shipping the full stack: the connection flow, backend data fetching/sync logic, storage, and the frontend analytics UI, plus the operational details like retries, rate limits, and graceful degradation when Apple data was delayed or unavailable.
Internal Tools
I built an internal tool that processes recorded sales calls and turns them into structured, queryable data for training. The pipeline ingested call metadata and transcripts, ran an LLM-based extraction step to produce normalized fields (summary, pain points, objections, competitor mentions, key moments), and stored the results so they could be searched and filtered.
On top of that dataset, I shipped a simple UI that lets reps browse calls by category, search for specific objections or topics, and open curated examples quickly. The main technical work was making the analysis repeatable and consistent (schema-first outputs, validation, and retries) and keeping it usable at scale (indexable storage and fast retrieval).
Reflection
I learned a lot. I did a lot. And somewhere in the middle of it, I figured out that a startup is exactly where I want to be.
I like being somewhere the work is real and the pace is fast. A startup is that place.
Huge shoutout to Anton Reza, Matt Espinoza, and Eshan Betrabet for taking a chance and bringing me on.