Seeking an experienced developer or team to create two distinct, yet integrated, automated systems designed to streamline business operations and enhance data management within my Notion CRM. I'd like to break down both systems into two phases with phase 1 focused on the core build (MVP) and phase 2 focused on enhancements.
Ideally I would like someone to both phase 1's but - if your expertise is primarily in one system (Audio Capture or Article Extraction), please indicate that in your proposal — we may split the work if needed.
The first system is an Audio Capture and Follow-Up Trigger system:
**What You'll Build**
An end-to-end automation pipeline:
-
Fireflies.ai → Auto-records & transcribes scheduled calls + impromptu calls (manually records)
-
Make.com → Orchestrates the workflow, handles matching logic & error recovery
- ChatGPT API → Converts transcripts into executive summaries + extracts structured metadata (deal types, markets, action items)
- Notion CRM → Receives clean, timestamped interaction records that also auto-triggers follow-up sequences
Core Features:
- Participant auto-matching (calls → correct client/partner profiles)
- Idempotency checks (no duplicate entries)
- "Needs Review" holding area for low-confidence matches
- Auto-tagging system (topics, deal structures, markets, asset types)
- Error handling with retry logic + Gmail alerts
- Weekly digest summarizing all conversations, trending topics, and upcoming follow-ups
**Requirements**
Must-Have Experience:
- Advanced
Make.com (complex workflows, webhooks, error handlers, routers)
- Notion API (databases, relations, formulas, bulk operations)
-
Fireflies.ai integration (webhooks, auto-join settings)
- Openai api (gpt-4 prompt engineering for structured outputs)
- data matching logic (fuzzy matching, confidence scoring)
- idempotency patterns (prevent duplicate writes)
---
deliverables:
- fully functional
make.com scenario(s) with all modules configured
- Notion workspace structure (4 databases: People, Interactions, Needs Review, Automation Log)
- Custom GPT prompts optimized for executive summaries + metadata extraction
- Error handling & retry logic with alerting
- Testing documentation proving 95%+ success rate across 20+ test calls
- Handoff video walkthrough (30 min) showing how to operate, troubleshoot, and maintain the system
- Written playbook (troubleshooting guide, monthly maintenance checklist)
Success Metrics:
- 95%+ automation success rate (calls logged without manual intervention)
- <5 min average processing time (Fireflies webhook → Notion write)
- <5% error rate with graceful fallbacks
- 90%+ participant match accuracy
- System runs hands-free for 30 days post-launch without breaking
---
Budget & Timeline
- Budget: Fixed price (quote your rate based on scope)
- Timeline: 2-3 weeks preferred
- Payment structure: 50% upfront | 50% on successful delivery & testing
-------
The second system is an Article Scraping system:
This system automatically scans two select CRE publications — to identify and extract mentions of equity backers involved in development or acquisition transactions. It filters out irrelevant content and isolates key deal intelligence focused on the equity partner when mentioned & a concise one-line summary of the transaction/partnership. Each extracted entry is pushed into the Equity database in Notion, which links directly to the Firms database & company profile in the CRM through a single property – Mentions/Deals (relation) - allowing multiple article mentions to attach to the same firm without ever creating duplicate company profiles, all within a single unified database.
Pre-filter keywords for articles before ChatGPT (“equity”, “joint venture”, “recap”, “financing”, “backed by”, “partnered with”). Cuts cost/noise.
In the Equity database we’ll have the following extracted from the article:
Title (Deal Headline / Summary)
- Format: [Equity Partner] + [Developer] – [Structure]
- *Example: “PGIM + Middleburg Communities – Platform Equity”*
Properties:
- Date Published
- Firm Category (relation to Firms database)
- Structure (JV Equity / LP Equity / Platform Equity / Recap / Co-GP / Seed Equity / etc )
- Market (State)
- Developer (relation to Firms database)
- Equity Partner (relation to Firms database)
- Source (The Promote, The Real Deal)
- Transaction Summary (one-liner summary)
- Confidence Score
- Article ID (Text used for idempotency, e.g., Hash of url + date + equity + developer)
- Article URL
- Processed At (created time for audit and weekly digests)
---
- Technical depth: Combines ai, workflow automation, data matching, and crm architecture
- portfolio piece: this is a showcase-worthy system for any automation specialist
this is a contract position for a one-time build + 30-day support period. Potential for ongoing retainer if system requires scaling/enhancements.
Prazo de Entrega: Não estabelecido