Rajat V.

Fractional AI CTO, and a sitting CTO today. For 4+ years I've led technology at a US PropTech AI startup, and the role goes well beyond code: I'm part of its investor meetings, pitch decks and fundraising demos. I also run my own services company, so I know sales and revenue firsthand. ROLES I FILL Fractional CTO, interim CTO or technical advisor, for founders who need a CTO or technical co-founder before they can hire one full-time. I also work forward-deployed, the way a Forward Deployed Engineer (FDE) does: embedded in your team, on your data and systems, until it's live and earning. BEYOND THE CODE - Investors: I know what investors ask a CTO because I've been in those meetings. I get the technical side of a raise ready: the tech story in your pitch deck, the demo, and technical due diligence. - Go-to-market (GTM) and revenue: I run 31Web LLC end to end (sales, pricing, hiring and delivery), and I track which features earn how much, so roadmap choices are weighed against revenue. - Growth and ads: working knowledge of performance marketing, including Google and Meta ads, ROAS and CAC, attribution and the Meta Conversions API (CAPI). I'm not your media buyer, but I make sure what we build and track matches what your ads pay for and what your sales team closes. - Leadership: 8+ years in leadership roles at US companies, hiring and leading engineering teams. - Decisions with consequences: architecture, model and vendor choices, build or buy, AI running costs and cloud FinOps, made by someone who lives with them every week. Code can be learned from a course or an AI tool. Knowing what investors ask, what sells and what a click is worth only comes from doing it. WHERE I HELP - PropTech and real estate: MLS-grade AI image generation, and AI that replaces manual design and staging work. - Legal tech and law firms: retrieval-augmented generation (RAG) over case files and documents, intake automation, and secure multi-tenant hosting. - Agentic AI: AI agents and agentic workflows that do real work in your systems, with human-in-the-loop (HITL) checks, and Model Context Protocol (MCP) servers that give agents safe access to your data. - AI in production: AI integration into the CRM, phone, payment and document systems you already run; custom-trained and fine-tuned models; LLMOps with evals, guardrails and observability. - Voice AI and AI receptionist agents: intake and speed-to-lead agents that answer fast, stay on approved facts and book appointments in GoHighLevel (GHL) or your CRM, with A2P 10DLC registration done right. - MVP to product-market fit (PMF): what to build first, what to cut, and taking AI-built (vibe-coded) or Lovable prototypes to production-grade code. - CRM marketplace apps: Salesforce AppExchange (ISV, managed packages), GoHighLevel and Pipedrive Marketplace apps that pass review and stay compliant. RESULTS - MLS-grade AI imagery: we trained our models to produce staged photos that meet MLS rules. Furniture changes; walls, windows and camera angle stay true to the property. Failed renders fell from over 20% to about 10%. - AI in place of manual work: AI now does the design and staging work human designers used to do by hand, for real-estate agents and for flooring and furniture retailers. - Voice AI: as CTO of a conversational-AI startup, I took its GoHighLevel voice agents from idea to a working MVP, and my team built the AI phone agent for a field-service platform used by contractors. Both run with low latency and hallucination guards. - Legal tech: ran a multi-tenant legal-tech platform on AWS from 2021 to 2024. - CRM platforms: built HighLevel's drag-and-drop page builder in its early years (2019 to 2021), reporting to its co-founders Shaun and Varun. Since 2024 I've built and maintained one platform's Salesforce, GoHighLevel and Pipedrive apps. HOW I WORK - I stay after launch: monitoring, evals, retraining and vendor changes are part of the engagement. - Decisions come in writing, including when to buy a tool instead of building one. - My team at 31Web LLC ships with AI-assisted engineering (Claude Code and agentic coding), under my review. - Flexible hours, with overlap across US time zones. - Several clients have kept working with me for 3 to 5 years. - My CTO role is a direct contract, off Upwork, so most of my Upwork history is hands-on builds at developer rates. This rate is for CTO-level work. STACK - Cloud: AWS, Google Cloud (GCP), DigitalOcean, Railway, Replit, Docker - AI: OpenAI, Gemini, Hugging Face, Replicate, fine-tuning, RAG, vector databases, MCP, ElevenLabs, Retell AI, Twilio - Apps and data: Python, Node.js, TypeScript, React, Next.js, Laravel/PHP, PostgreSQL, MySQL, MongoDB - Growth: Google Ads, Meta Ads, GA4, Meta CAPI, GoHighLevel and CRM pipelines START HERE Book a 30-minute consultation, or start with a fixed-price AI Readiness & Architecture Review. Send your workflow, spec or repo, and my reply will open with the three questions I'd ask first.