Jump to projects, pages, or hiring context.
Senior Software Engineer at Toosi · AI Software Engineer · Bengaluru
7+ years shipping production ML, GenAI and full-stack systems. Most recently I built Toosi, an autonomous outbound platform with eval gates, human approvals and billing driven by an LLM classifier, from idea to 2 paying clients with live campaigns across the US, UK and Ireland.
Open to Senior AI Software Engineer roles · Remote (US / UK / EU overlap) or hybrid Bengaluru
7+
Years
Building production software & AI
Idea → Revenue
Toosi, 0 → 1
2 paying clients, campaigns live in US, UK & Ireland
1K+
Concurrent requests
ML services on Azure Kubernetes at Rubick.ai
79%
Prediction accuracy
Next-purchase model, ~10% AOV uplift at Krowd
Give me an objective, a messy workflow and access to users, and I start thinking about the intelligence layer, data model, architecture, product experience and how the system improves after deployment.
Today I'm a Senior Software Engineer at Toosi. Before that: Tech Lead on ML platforms and MLOps at Text Mercato Solutions (renamed Cataloging.ai, now Rubick.ai), and recommendation engines, a merchant intelligence graph and web intelligence at Krowd. On the side, Neo-AI, a voice-first agent you can talk to, interview with, learn from and practise a language with.
What interests me now: systems that understand context, reason over information, use tools, act autonomously and get better through feedback.
Sanjeet Kumar · Bengaluru
Systems with real ownership — architecture, intelligence, product and production. Not weekend demos.
Toosi TravelProductionAn autonomous outbound platform, built 0 to 1
ICP → discovery → enrichment → LLM-written outreach from the client's inbox → reply classification → billing only on positive replies.
Business impact
Took the product from idea to real revenue: 2 paying clients on a $50-per-positive-lead model, with live campaigns across the US, UK and Ireland, which I built and took to market end to end.
Rubick.ai (formerly Text Mercato)ProductionML platform, MLOps and cataloging infrastructure
MLflow experiments → containerised models on Azure Kubernetes → NestJS gateway with Auth0 → cataloging platform workflows → marketplace-ready catalogs.
Business impact
The cataloging platform became the daily tool for 70+ specialists, ML services scaled to 1K+ concurrent requests, and I led the technical onboarding of two clients, taking AI capabilities from experiment to working integration.
Toosi TravelProductionSelf-onboarding & Q&A agent
Prospect question → RAG-grounded answer → workflow-step check → progressive self-onboarding, with conversation state tracked throughout.
Business impact
Restaurant prospects now self-onboard from their first question without waiting for a sales call, so the sales team stopped being the bottleneck for sign-ups. Live in production at onboarding.krowd.ai.
IndependentVoice-first agent platform: talk, interview, tutor, learn a language
Speak → streaming STT → agent with tools (code sandbox, tests, whiteboard) → spoken reply in ~1s; a separate evaluator scores interview answers against an 18-competency rubric. Runs fully local.
Status
Pre-launch prototype: the spoken talk → think → speak loop, the four capabilities and the interview engine with 18 competencies, evidence-based scoring, a coding sandbox and a whiteboard run end to end on one Mac with zero inference cost and no audio leaving the machine. Next: browser widget UI, persistent memory, hosted beta.
Toosi TravelProductionTurning the open web into structured sales intelligence
Web discovery → enrichment → validation → entity resolution → structured leads for autonomous sales agents.
Business impact
Replaced manual list-building with continuous discovery and enrichment that feeds qualified leads directly to Toosi's agents: 40K+ entities per city today, with an architecture targeting 100K+ businesses and 300K+ contacts.
KrowdProductionPersonalized recommendations and transaction intelligence
Crawled reviews → NLP → MALLET topic vectors → user preference profiles → location-aware ranking → feedback loop; later transaction sequences → next-purchase prediction → offer matching.
Business impact
Became Krowd's first externally demonstrable personalization product, used by the co-founder in sales, investor and partnership conversations. The transaction-intelligence work reached ~79% next-purchase accuracy and lifted average order value by about 10% for partners including Barclays, HSBC, Amex and Dubai Islamic Bank.
Jul 2023 — Present
Toosi Travel Pvt Ltd
Senior Software Engineer at Toosi, building the company's technology end to end: architecture, agent orchestration, web intelligence, backend, full-stack product, evaluation and production deployment on AWS.
Apr 2021 — Jul 2023
Text Mercato Solutions Pvt Ltd · Renamed Cataloging.ai, now Rubick.ai
MLOps Engineer → Tech Lead
Built and operated production ML and backend platforms used directly by customers, across Kubernetes, API gateways, catalog intelligence, price tracking, CI/CD and full-stack product development.
Dec 2018 — Apr 2021
Krowd R&D India Pvt Ltd
Software Engineer Intern → SDE-1
Joined as the company's first intern and grew into an SDE-1 across recommendation systems, web intelligence, data infrastructure, Kafka, graph databases and cloud automation.
Organized around what I can build — not a logo cloud.
Recommendation engines, merchant graphs, web intelligence and ML platforms, all in production. The thread didn't start with ChatGPT.
KrowdProductionWent into production as the data layer behind Krowd's campaign platform: a Neo4j merchant → outlet graph of ~20K UK restaurants with incremental, entity-resolving ingestion, later grown to ~70K nodes across the UK, Ireland and the US.
Python · Neo4j · Entity resolution · Graph data modelling
Read my hiring brief for the 90-second version: proof points, stack, work history and how to reach me.