
Verify
People stay in the loop where a mistake is expensive
Every handwritten deed goes to a human reviewer, however confident the model is. On a legal document, slower beats confidently wrong.

Open to AI engineering roles in London, ideally in financial services or insurance.
AI engineer in London. I build LLM agents, document pipelines and voice systems that run in production.
Hi, I'm Vishit's AI. I answer from his verified project notes, so I won't make things up. Ask about anything he has built, paste a job description to see where he fits, or book a call with him.
What the AI looked up
Before answering, the AI checks his notes. Each check appears here.
It answers from my verified project notes, shows you every lookup it makes, and can book a call on my calendar.

Bank lawyers in India check a property's title by reading scanned land records. EasyHunt reads them for the lawyer and drafts the report. It takes up to 20 documents per case (7/12 extracts, mutation records, sale deeds), in several languages and some handwritten, and produces a scrutiny report in the bank's own template for the lawyer to review and approve.
Businesses create voice agents, give them a phone number and a knowledge base, and get the call logs. It is built in the style of Retell AI, and every customer's data is isolated in its own workspace.

A personalised live football broadcast, built for the Microsoft x Premier League hackathon. A team of AI agents decides which moments matter, explains them for each viewer in their language, and proves every number before it airs.

A live risk map of the Gulf, with situation reports and SafeConnect, a way for families to check in with each other. It reads news and feeds from 12 countries every 15 minutes and shows a risk level for each country on a 3D globe.

Builds AI agents and voice agents for businesses. The research side is about keeping a person's voice, memories and way of thinking after they pass away within reach of the family who loved them.
Voice and video interviews built around the role and the candidate's CV, with proctoring and a written evaluation for the hiring team.
Also built: Credit memo agent, Credit-risk signals from SEC filings, Crypto trading bot, Personality from speech, On-device voice assistant, Agentic HR, hiring workflow system, RAG chunking strategies, Muse Meet, NextRole, explainable resume matching, TeslaTrend AI.
I lead a small AI team at Encegen AI Labs, I'm the founder of Novessa, and I document my builds on YouTube.

Every handwritten deed goes to a human reviewer, however confident the model is. On a legal document, slower beats confidently wrong.

On a phone call a second of silence feels broken, so every stage streams instead of waiting for the one before it to finish.

The lending pipeline has 164 automated tests, and they run on every change.

This site rate-limits its agent, asks for a clear yes before it books anything, and stops it repeating its own instructions.
This site runs a public MCP server. Add it to Claude, Cursor or any MCP client, then ask your own assistant whether I fit your role. It reads the same verified notes as the agent above, and it can book a call too.
https://vishitjiwane.com/api/mcp
Open to AI engineering roles in London, ideally in financial services or insurance. Needs Skilled Worker sponsorship. Qualifies at the new-entrant rate, which lowers the salary threshold the role has to meet. My agent checks my calendar and sends you a calendar invite.