Real-time AI Voice Receptionist
Missed after-hours and overflow calls → captured bookings with sub-300ms voice responses that felt live.
About
Pranav Desai
AI Systems Architect · Forward Deployed Engineer
I deploy production AI agents for SMB customers and own delivery end to end, from discovery and success criteria through proof of concept to production. Voice and chat agents, RAG pipelines, and multi-agent sales-ops, built so teams actually adopt them.
Experience
A decade from infrastructure to AI delivery.
- ConsultOps.ai2023 - PresentFounder & Principal AI Solutions Architect
Founded and run the business end to end: find the customer, scope the problem, price it, build it, and support it in production.
- Shipped a real-time AI voice receptionist that turns after-hours and overflow calls into booked work.
- Built RAG illustration and onboarding automation that cut manual prep ~85% per client.
- Delivered multi-agent sales-ops, website chat agents, and internal RAG assistants across 8+ clients on a reusable agent framework.
- New Relic2022 - 2024Pre-Sales Solutions Consultant
Owned technical pre-sales for enterprise observability customers across the evaluation cycle.
- Built tailored proofs of concept and demos that turned complex telemetry into clear business outcomes.
- Partnered with engineering on data pipeline reliability and customer success criteria.
- Aptum Technologies2020 - 2022Pre-Sales Engineer
Designed cloud and resiliency solutions for regulated industries.
- Architected multi-cloud solutions across AWS, Azure, and DRaaS with compliance in mind.
- Automated proposal and RFP workflows to speed the pre-sales cycle.
- Hewlett Packard Enterprise2016 - 2020Sr. Solutions Consultant
Enterprise storage and access-control consulting for large accounts, plus technical workshops.
- Reliant Technology2014 - 2016Storage Administrator
Storage operations and DoD-7 compliant monitoring for secure environments.
Stack
What I build with.
Agents & LLMsAnthropic Claude, OpenAI, OpenRouter, multi-agent systems
Orchestration & AutomationClaude Code, MCP, agentic workflows, RAG, AIOS, REST APIs, CI/CD
LanguagesPython (FastAPI), TypeScript, SQL
Cloud & EdgeAWS, Azure, Cloudflare (Pages, Access, DNS)
DataPinecone, Google Cloud APIs
Web & DiscoveryAstro 7, SEO / AEO / GEO, schema, Google Business Profile, GA4, Search Console, Cloudflare Web Analytics, Lighthouse
DesignFigma, Claude Design, Canva, Higgsfield
CertifiedAWS Solutions Architect, AssociateAzure Solutions Architect
EducationB.S. Computer Engineering, Kennesaw State UniversityA.S. Electrical & Computer Engineering, Savannah Technical College
Use cases
Systems running in production.
Service businessAI Voice Receptionist
Insurance / financeRAG Onboarding Automation
Marketing agencyLead-Gen + CRM Orchestration
B2B SaaSInternal RAG Assistant
B2B distributorMulti-Agent Sales-Ops
E-commerceWebsite Chat Agent
Professional servicesDocument Extraction Pipeline
ConsultOpsReusable Agent Framework
Twenty-one production systems across eight clients. Each one is real, anonymized by industry, which is standard under NDA.
RAG Illustration + Onboarding Automation
Hours of hand-built illustrations and paperwork → roughly 85% less manual prep per client.
Lead-Gen + CRM Orchestration
Leads slipping through the cracks → near-instant response, with the team working only qualified leads.
Internal RAG Knowledge Assistant
Tribal knowledge buried in docs → fewer repeat questions and faster onboarding for new hires.
Multi-Agent Sales-Ops System
Reps buried in research and drafting → time back for real conversations, not busywork.
Website Chat Agent + Backend Integration
Shoppers asking what a static site could not answer → higher engagement and qualified leads off the site.
Document Processing + Extraction Pipeline
Slow, error-prone manual review → roughly 85% less review, with humans on exceptions only.
Reusable Prompt + Agent Framework
Every build starting from scratch → faster, more consistent deployments across clients.