Fractional AI CTO for companies
building serious AI products.
I help founders take AI products from MVP to reliable production systems — across RAG, Voice AI, Agents, Search and enterprise AI.
Production systems, not demos
A selection of AI systems I’ve architected and shipped — from venture-backed GTM AI to multilingual search at national scale.
Sprouts.ai
AI engineering for a venture-backed Revenue AI platform used by Hewlett Packard, Razorpay, HighRadius and Udemy — target-profile generation, enrichment, AI workflows and GTM automation.
Openclaw
Architected and scaled autonomous multi-agent workflows powering the product’s core operations.
Commonplace Forum
Shipped the governance layer securing multi-agent AI for a U.S. Department of State education trial.
Enterprise voice deployments
Deployed sub-800ms real-time AI voice agents for enterprise clients, replacing manual and legacy IVR-driven workflows.
Unhamster
Built a voice/video agent platform from scratch, including self-hosted LLM infrastructure, with 500+ UAT users and multiple live client deployments.
Jolii
1,000+ beta users. LangChain-based contextual learning and search built around a YouTube learning pipeline.
Jio / Reverie
Multilingual semantic search across Indian languages at production scale — experience that predates the current LLM boom.
Voice Eval
A provider-neutral simulation and grading module for voice AI pipelines — scripted multi-turn conversations, per-turn audio capture, rule-based grading and cross-pipeline comparison, run before every deploy instead of manual retesting.
dsh-minimal-anchor
A DeepSeek Harness plugin that prunes a coding agent’s tool schema on a session’s first turn only, keeping early reasoning on exploration instead of premature action — then gets out of the way from turn 2 onward.
Nudgit Lab
Technical advisor to an AI companion/coaching platform running real-time voice conversations through LiveKit and Pipecat, with swappable STT/LLM/TTS providers and a Gemini Live speech-to-speech path.
Where AI products actually break
AI architecture
Production reliability
RAG / AI search
Voice AI
Agentic AI
Evaluation
AI infrastructure
Engineering leadership
Check before you hire anyone
60-second, rule-based diagnostics — not a lead form dressed up as a quiz.
AI Visibility Check
See how discoverable, understandable and citation-ready your website is across search and AI answer engines — scored from the live page, not a guess.
Does My Startup Need a Fractional AI CTO?
5 questions. Find out whether you need a senior AI engineer, a technical advisor, a Fractional AI CTO, or a full-time hire — right now, not eventually.
Four ways to work with me
AI Product & Architecture Assessment
A structured review of your architecture, RAG/retrieval, agents, voice, evaluation, security, reliability and cost — delivered as an AI Production Readiness Score with prioritized risks.
Fractional AI CTO
Ongoing technical leadership: architecture decisions, AI roadmap, hiring, vendor/model decisions, evaluation strategy and production reliability.
AI Technical Due Diligence
For investors, accelerators and acquirers who need an independent technical read on an AI startup before committing capital.
AI Architecture Consulting
Focused engagements on RAG, voice AI, agentic systems or AI search — for teams that need deep expertise on one system, not a full-time CTO.
Start where you are
AI Product Health Call
Free · 20 minutes · qualificationAI Product & Architecture Assessment
2 weeks · written report & readiness scoreFractional AI CTO
Advisory to deep engagement · monthly retainerEngineering execution via Jalpa
Build & ship the roadmapI’ve been building production AI since before “AI startup” became a category
2017
Conversational AI and recommendation systems.
2018–2021
Enterprise conversational AI, NLU and knowledge graphs / search.
2022–2023
Multilingual search at production scale (Jio / Reverie).
2023–2026
LLMs, RAG, agents, voice AI and AI SaaS engineering.
Today I work as a Fractional AI CTO for founders building AI products, and lead Jalpa, a production AI engineering studio, when a roadmap needs a team to execute it. I’m also Co-founder & CTO of Rasadhara AI Labs, where I hold direct technical leadership responsibility rather than just advising on it.
Verified feedback, not selected quotes
Pulled directly from completed, rated Upwork engagements.
“Abhishek delivered production-ready AI infrastructure for our platform, integrating multiple LLMs (GPT-4, Claude), vector search with Qdrant, memory persistence, and agent orchestration. His backend work using FastAPI, LangChain and tracing systems was top-tier. Highly recommended for complex AI/LLM systems.”
“I had a great experience with Abhishek. He completed the project within the time frame and helped me understand the changes, which was very helpful. I would like to recommend him for my future projects.”
“Abhishek was wonderful. He made the process of setting up and integrating Ollama with Fabric very easy. He even dedicated his own time to make sure the setup session was efficient. I would definitely work with him again.”
What is a Fractional AI CTO?
A Fractional AI CTO is a senior technology leader who helps companies make AI architecture, product, engineering and infrastructure decisions without hiring a full-time CTO. Unlike a general fractional CTO, the role requires direct experience with production LLM systems — evaluation, retrieval, agents, model infrastructure and AI reliability.
Need a conversation, not a pitch.
30 minutes. Tell me what you’re building and where it’s breaking.