About Me
Founder and product-builder focused on generative AI / LLM products and data-driven product strategy, with hands-on experience shipping RAG systems, prompt-less LLMs, and consumer/education pilots.
Key founder projects
…
Founder and product-builder focused on generative AI / LLM products and data-driven product strategy, with hands-on experience shipping RAG systems, prompt-less LLMs, and consumer/education pilots.
Key founder projects
Raavann — Deep Research LLM Engine (Founder, Jun–Nov 2025): Built a prompt-less research LLM that recommends and can remotely execute experiments via serverless GPU infrastructure; piloted with universities and launched across London, Coventry, Belfast and India.
Zaplet AI — Rental Property Platform (Founder, Sep 2024–Jan 2025): Aggregated ~50k London listings (major agencies), built a RAG LLM enriched with cost-of-living/amenities/crime context, and an interactive Tube-map tap-to-query UX; validated map-driven flows via user testing.
Personalised Productivity Suite (Feb–Jun 2025): Built AI-native Word/PowerPoint/Excel + diagram generator; features include auto-structuring, table/formula reasoning and prompt-less diagramming — student pilots showed ~2.5× faster outcomes.
Professional experience
Head of Operations & Technology — Corker Ltd, London (Oct 2023–Sep 2024): Led integrations and automation (Zapier/APIs), increased monthly physical outreach from 50 → 1,650 (≈3200% increase) with a 23% response rate; migrated systems to Odoo (6 months) improving efficiency ~40%; automated inventory updates (saved ~100 hours/month) and deployed GPT-based customer automation cutting response time 24h → <2h.
Founder — Vector Motors, Mumbai (Apr 2020–May 2021): Raised £50k, built supply chain across Asia, reduced lead times 25%, and managed design/production of an electric motorcycle prototype.
Education
Warwick Business School, MSc — Management of Information Systems & Digital Innovation (Sept 2022–Sept 2023). Top team project score (85%); dissertation analysing 150 startups to identify factors behind unicorn outcomes.
KIIT, B.Eng (IT)
Experience
Founder
<p>Built world’s 1st Deep research engine exclusively for AI research.<br>
Designed world’s 1st research LLM that operates without prompts.<br>
Designed a tool that recommends which experiments to run.<br>
Integrated serverless GPU infrastructure to execute experiments remotely from anywhere.<br>
Conducted offline pilot tests at universities.<br>
Launched across London, Coventry, Belfast, and India.<br>
Used custom multi agent architecture.<br>
Used knowledge graphs.<br>
Used vector embeddings.<br>
Used semantic search.<br>
Used prompt engineering.<br>
Used context engineering.<br>
Used ChatGPT API integrations.<br>
Used PR driven pull product.</p>
Founder
<p>Designed and built AI-native Word, PowerPoint, Excel, and an architecture-diagram generator from the ground up.<br>
Drove development via continuous user feedback with pilots with engineering students at the University of Warwick and KIIT.<br>
Drove development via research-led improvements.<br>
Shipped automatic document/slide structuring.<br>
Shipped table reasoning with formula suggestions.<br>
Shipped slide narration.<br>
Shipped prompt-less diagramming from plain-language specs.<br>
Achieved sustained student adoption across coursework/projects.<br>
Students achieved their goals 2.5x faster than alternatives available.</p>
Founder
<p>Built an AI-driven London rental search platform.<br>
Aggregated approximately 50k listings from Savills, Knight Frank, Winkworth, and Hamptons.<br>
Excluded SpareRoom due to map-exposed addresses.<br>
Implemented a RAG LLM enriched with cost of living, amenities, crime type + location, parks, hospitals, and gyms for contextual answers and comparisons.<br>
Created a fully controlled, interactive Tube map using station coordinates with POI search.<br>
Enabled tap-to-query flows such as within 20 minutes’ walk of Aldgate.<br>
Ran regular user tests.<br>
Validated strong preference for map-driven selection over free-form prompts.<br>
Iterated UX accordingly.<br>
Concluded that innovation depends on controllable components.<br>
Concluded that consumer rental search is a low-frequency usage market, making heavy innovation cost-inefficient at scale.</p>
Head of Operations and Technology
<p>Integrated Searchland, Excel, Docmail, Keap, and ClickUp through Zapier and API integrations.<br>
Increased monthly client outreach from 50 to 1,650 physical letters.<br>
Achieved a 3200% increase in outreach.<br>
Achieved a 23% response rate.<br>
Led the transition from Intact IQ to Odoo.<br>
Completed the project in 6 months.<br>
Enhanced operational efficiency by 40%.<br>
Developed an automated system to update product information from supplier emails to the company website using WooCommerce.<br>
Improved inventory management accuracy by 35%.<br>
Saved 100 labor hours monthly.<br>
Automated responses to basic client inquiries using GPT integrated with Zapier.<br>
Reduced response times from 24 hours to under 2 hours.<br>
Handled over 500 queries per month.<br>
Conducted comprehensive qualitative research to prioritize improvements for over 20 business processes.<br>
Resulted in a 30% increase in operational efficiency.<br>
Created a centralized repository for 500+ documents.</p>
Head of Operations and Technology
Integrated Searchland, Excel, Docmail, Keap, and ClickUp through Zapier and API integrations, increasing monthly client outreach from 50 to 1,650 physical letters (a 3200% increase) and achieving a 23% response rate. Led the transition from Intact IQ to Odoo, completing the project in 6 months, enhancing operational efficiency by 40%. Developed an automated system to update product information from supplier emails to the company website using WooCommerce, improving inventory management accuracy by 35% and saving 100 labor hours monthly. Automated responses to basic client inquiries using GPT integrated with Zapier, reducing response times from 24 hours to under 2 hours and handling over 500 queries per month. Conducted comprehensive qualitative research to prioritize improvements for over 20 business processes, resulting in a 30% increase in operational efficiency and creating a centralized repository for 500+ documents
Founder
<p>Secured £50,000 in startup funding.<br>
Designed a diverse team development plan.<br>
Established international supply chain networks across Asia.<br>
Reduced delivery lead times by 25%.<br>
Managed the design and construction of the first electric motorcycle prototype.<br>
Incorporated insights from BMW's electric vehicle division.</p>
Founder
Secured £50,000 in startup funding, designed a diverse team development plan, and established international supply chain networks across Asia, reducing delivery lead times by 25%. Managed the design and construction of the first electric motorcycle prototype, incorporating insights from BMW's electric vehicle division
PROJECTS
House Hunt
<p>Raavann — a deep-research LLM (Jun–Nov 2025). Built a prompt-less research engine tailored for AI research: automates literature synthesis, precise search, and insight extraction. Validated in university trials and pilot launches across London, Coventry, Belfast and India to boost researcher productivity.</p>