Modular AI Workflow Development for Service Business
Budget: $30 – $250 USD
AI Workflow Specialist Needed – Modular Phone + Email + Automation Tools for Small Service Business (Open-Source Focus)We are a small service business looking for an experienced AI freelancer to help us build practical, modular workflow tools in low-risk phases. We are not looking for one large all-in-one rebuild. We want to test a few high-value tools first, see clear results, and expand only after success.Our strong preferences Fully modular, easy-to-extend, and self-owned systems
Strong emphasis on open-source tools, especially Qwen-Agent, Qwen-based models, DeepSeek-based workflows, or similar
Build fast, avoid unnecessary complexity, use proven open-source components
Everything must be designed so we can keep adding new data, rules, prompts, and workflows later without rebuilding
Someone who can coach/guide us — we have zero experience with agents
Priority Tools (to be built in phases)Unified AI Phone Agent (Intake + Pricing & Availability)
A single professional inbound call handler that: Greets callers politely and explains the owner is currently on another job
Collects name, callback number, service address, job details, urgency level, and best contact method
Answers basic pricing and availability questions using only our pre-approved rules and pricing sheets
Offers callback or routes to email when appropriate
Never makes custom quotes or promises outside the rules
Logs every interaction and saves structured intake data for later review
Email Text Extraction Workflow
Reads customer email threads and converts the text into structured data (name, phone, address, service requested, timing, summary). Flags missing information if we need to follow up.
Note: No current AI (including Gemini Workspace) does this reliably well for our needs.
Automated Confirmations & Reminders
Sends job confirmations and timed reminders (3 days, 2 days, 1 day before) via email and/or text. Logs all sent messages.
Outgoing Problem / Damage Intake Agent
Makes outbound calls for reported issues: apologizes, listens, gathers facts (date, location, service performed, damage details, photos/video request), logs everything structured, and routes for owner review — without making promises or liability statements.
What We’re Looking For Deep experience with Qwen-Agent / Qwen models and DeepSeek
Experience building in small, phased deliveries
Strong coaching mindset for non-technical clients
Ability to recommend a clean, open-source stack (don’t overbuild)
Proven use of modular frameworks so the system stays fully under our control and flexible long-term
Please reply with: Relevant examples of similar tools/agents you’ve built using Qwen or DeepSeek (especially voice/phone agents and email parsing)
What you specialize in
Your suggested tech stack + why (must be open-source friendly)
Estimated timeline for Phase 1 (Unified Phone Agent)
First-phase pricing
How you would design the system so we can keep adding rules, prompts, and new workflows easily while keeping full ownership
If you have strong experience with Qwen-Agent, DeepSeek, and open-source voice frameworks, that is exactly what we are looking for.If this sounds like a good fit, send your experience and approach. We’re ready to start with a small first phase.Thank you!
Strong emphasis on open-source tools, especially Qwen-Agent, Qwen-based models, DeepSeek-based workflows, or similar
Build fast, avoid unnecessary complexity, use proven open-source components
Everything must be designed so we can keep adding new data, rules, prompts, and workflows later without rebuilding
Someone who can coach/guide us — we have zero experience with agents
Priority Tools (to be built in phases)Unified AI Phone Agent (Intake + Pricing & Availability)
A single professional inbound call handler that: Greets callers politely and explains the owner is currently on another job
Collects name, callback number, service address, job details, urgency level, and best contact method
Answers basic pricing and availability questions using only our pre-approved rules and pricing sheets
Offers callback or routes to email when appropriate
Never makes custom quotes or promises outside the rules
Logs every interaction and saves structured intake data for later review
Email Text Extraction Workflow
Reads customer email threads and converts the text into structured data (name, phone, address, service requested, timing, summary). Flags missing information if we need to follow up.
Note: No current AI (including Gemini Workspace) does this reliably well for our needs.
Automated Confirmations & Reminders
Sends job confirmations and timed reminders (3 days, 2 days, 1 day before) via email and/or text. Logs all sent messages.
Outgoing Problem / Damage Intake Agent
Makes outbound calls for reported issues: apologizes, listens, gathers facts (date, location, service performed, damage details, photos/video request), logs everything structured, and routes for owner review — without making promises or liability statements.
What We’re Looking For Deep experience with Qwen-Agent / Qwen models and DeepSeek
Experience building in small, phased deliveries
Strong coaching mindset for non-technical clients
Ability to recommend a clean, open-source stack (don’t overbuild)
Proven use of modular frameworks so the system stays fully under our control and flexible long-term
Please reply with: Relevant examples of similar tools/agents you’ve built using Qwen or DeepSeek (especially voice/phone agents and email parsing)
What you specialize in
Your suggested tech stack + why (must be open-source friendly)
Estimated timeline for Phase 1 (Unified Phone Agent)
First-phase pricing
How you would design the system so we can keep adding rules, prompts, and new workflows easily while keeping full ownership
If you have strong experience with Qwen-Agent, DeepSeek, and open-source voice frameworks, that is exactly what we are looking for.If this sounds like a good fit, send your experience and approach. We’re ready to start with a small first phase.Thank you!