Aeruku

Pitch Deck
Architecture
Temporal

Inspiration

We were inspired by the vision of democratizing precision customer acquisition - making it accessible to businesses regardless of size or budget. Temporal AI emerged as the perfect tool, allowing us to transform this challenge into a reproducible, workflow-driven solution that adapts and scales automatically.

What it does

Aeruku is an AI-powered Go-To-Market engine that automates the entire customer acquisition pipeline. Built on Temporal AI's workflow orchestration, our platform identifies high-intent prospects, enriches their profiles with actionable context, generates bespoke personalized messaging, and executes multi-channel outreach campaigns - all while continuously learning and optimizing for better conversion rates.

How we built it

We used TRAE for AI assistance with coding, and Temporal's AI Agent demo repo. Aeruku uses advanced workflow orchestration for four agents on Temporal each of which provides its own unique output. The result is a robust, scalable system comprising of Agent to Agent communication from MCP servers. The MCP servers are from Apify, the LinkedIn scraper which works well, in addition to the Arcade.dev which helped with Gmail integration. We currently use in-story memory so chat history is not retained. However, the resulting email drafts are indeed persistent on the connected Gmail account. MCP Protocol Stack:

Apollo integration for lead sourcing with rate limiting LinkedIn scraping through Apify's MCP server OpenAI text generation with context injection Arcade.dev email automation with deliverability controls

Challenges we ran into

We ran into the persistence of Temporal as problematic during development as we needed to keep refreshing. Additionally, activity timeout defaults were not helpful. We also encountered that TRAE sometimes takes longer with larger contexts. Aggregating real-time data from disparate sources while maintaining quality was also difficult.

Accomplishments that we're proud of

Fully Temporal-Native Architecture: Every core function runs as a Temporal workflow Scalability: This product can be used in a scalable and sustainable way by many users in the future

What we learned

Temporal is very fault tolerant for scaling our apps in the future Context is king in AI Sales Teamwork is everything and NEVER GIVE UP!

What's next for Aeruku

Connect it to a PostGRES server like Supabase Deploy it online Keep refining it for specific users and show the outputs instead of keeping it on the Temporal server