Selected Outcomes
Lead engineer on 3 production AI products: Kshetra.ai, AI SDR, AI Receptionist
Keep scrolling→Delhi — ISTLinkedIn↗
Lakshya Verma
Scoped · Built · Shipped · SignedCTO, Legit Ventures & Legit.Global, Fractional CTO for AI Startups, Delhi NCR, India,
I scope the product, choose the architecture, build the MVP, and get it live with paying users.
10Years building for the web
500+Paying customers, Kshetra.ai
3Production AI products
Worked withLegit VenturesLegit.GlobalPassionfrutSaltnewsWealthchoice InfratechKshetra.aiWeAssembleAAFT
(01) ProfileSince 2016
I’m a hands-on startup CTO with 10 years of building for the web and the last two years spent shipping production AI products. I led the build of Kshetra.ai, a vertical AI SaaS that grew past 500 paying customers.
Now
As CTO at Legit Ventures and Legit.Global, I lead technical strategy for Legit Ventures’ portfolio startups and engineering for Legit’s clients, such as Saltnews and Wealthchoice Infratech.
(01) The full arcIdea → Handover
My work covers the full arc.
Scope the product
I work with you to find the one workflow customers will pay for, and cut everything else from v1.
Scope docUser flowsv1 cut listChoose the architecture
I pick the stack, models and vendors on cost, speed and risk, and write down why, so the next engineer doesn't have to guess.
Architecture docBuild-vs-buy callsVendor picksBuild the MVP
I build the core workflow production-grade, with evals and guardrails on the AI parts and payments wired in from day one.
Production MVPEvals & guardrailsCheckoutLive with paying users
I launch, onboard the first customers, and watch real usage to fix what breaks and cut what nobody uses. That's how Kshetra.ai passed 500 paying customers.
LaunchFirst customersMonitoringStandards, team & docs
I set code standards, CI/CD and observability, plan the first engineering hires, and record every architecture decision.
CI/CDHiring planDecision recordsScale, or hand over
You keep me on to scale, or I hand over to a permanent CTO with the docs, access and context they need from day one.
Scale roadmapHandover packCTO onboarding
(01) My edge02 products · custom builds
Two AI products I built and run in production, plus custom flows for everything else:
Calling + WhatsApp AI SDR
(01)One sales rep on both channels. It calls and messages leads, qualifies them, and hands the ready ones to your team.
Page-aware AI receptionist
(02)It knows which page a visitor is on and answers for that page. It captures the lead and sends a WhatsApp with their project details, so the first conversation starts hot.
Custom AI workflows
(03)n8n and LLM flows built around how your team already sells, supports, and operates.
Each is built with evals, guardrails, monitoring, and a human in the loop wherever a wrong answer costs money.
(02) Services05 engagements
What I Do for Startups
(01) · Engagement01 / 05
Fractional CTO
I own the technical roadmap, architecture decisions, stack, and vendor choices. I also handle hiring plans for the first engineers and represent the technology to investors and the board.
- Technical roadmap
- Architecture
- Stack & vendors
- First engineers
- Investors & board
(02) · Engagement02 / 05
AI MVP Sprint
This is a fixed-scope build that takes a founder from idea to a launched product in weeks. It runs in three phases:
- DefineDefine the architecture and the one core workflow the product must do well.
- BuildBuild it production-grade, with evals, guardrails, and payment checkout.
- LaunchLaunch it and onboard the first users.
(03) · Engagement03 / 05
AI Automation for Growth & Ops
I build:
- WhatsApp lead qualification
- Inbound and outbound voice agents
- Chat agents
- CRM routing
- n8n workflows that remove manual work from sales, support, and operations
(04) · Engagement04 / 05
Tech Audit & Investor Readiness
I review the codebase, architecture, and infrastructure, then write a stabilization roadmap. I also prepare documentation and talking points for technical due diligence.
- Codebase review
- Infrastructure
- Stabilization roadmap
- Due diligence
(05) · Engagement05 / 05
Team & Process Setup
I set code standards, CI/CD and deployment processes, and observability. I document architecture decisions so the setup outlasts my engagement.
- Code standards
- CI/CD
- Observability
- Architecture decisions
(03) Selected Outcomes01 / 03
(01) Vertical AI SaaSkshetra.ai ↗
Kshetra.ai
500+Paying customers
I was the lead engineer from zero to revenue on this vertical AI SaaS. It reads an uploaded floor plan and generates a detailed Maha Vaastu report. I built the vision and LLM pipeline, frontend, billing, and deployment.
Vision + LLM pipelineFrontendBillingDeployment
(02) AI sales rep on calls and WhatsApp
AI SDR
~80%Of repetitive lead qualification, automated
One agent qualifies leads on WhatsApp and on live calls. It automates about 80% of repetitive qualification and replies in under a second on calls, handling interruptions. High-stakes replies go to a human for approval, and outcomes are written back to the CRM.
WhatsApp Business APIVoice callingLead scoringHuman handoffCRM write-back
AI SDR — WhatsApp lead qualification and a live call handing off to CRM
(03) Page-aware website agent
AI Receptionist
60%More leads convert from the website
It reads the page a visitor is on and answers for that page. When the visitor shows intent, it captures their details and sends a WhatsApp message with the project summary, so sales opens on a hot lead.
Page contextLead captureWhatsApp follow-up
AI Receptionist — page-aware chat on a website
(04) Production disciplineStage 01 / 07
Production discipline
Every build ships with retries, circuit breakers, prompt versioning, eval harnesses, observability, and runbooks. These patterns come from my open-source n8n-BMAD framework.
Leads arrive on the WhatsApp Business API, a Vapi or Twilio voice line, or chat.
- 01InboundWhatsApp · Voice · Chat
- 02Versioned promptPrompt versioning
- 03Model callRetries · Circuit breakers
- 04GuardrailsOutput checks
- 05Eval harnessKPIs as eval metrics
- 06Human in the loopApproval on high stakes
- 07CRM write-backOutcomes logged
Observability
Runbooks
(06) Experience2016 — Present
Experience
From2026Chief Technology Officer, Legit Ventures
Chief Technology Officer
- —I lead technical strategy, architecture, and engineering across Legit Ventures and its portfolio startups, turning unvalidated ideas into revenue-generating products in weeks.
- —I own the technical roadmap, build-vs-buy calls, and vendor evaluations, and prepare the architecture docs and technical due diligence investors ask for.
- —I set the engineering bar: CI/CD, latency and cost budgets, observability, circuit breakers, and runbooks.
Chief Technology Officer
- —I lead engineering for Legit and its startup clients, including Saltnews and Wealthchoice Infratech. My scope covers websites, web and mobile apps, AI automation, and the tracking behind performance marketing and SEO.
- —I run Legit’s AI MVP sprint and fractional CTO offering for founders.
- —I set the stack, code standards, and deployment process for every client build.
Full Stack Developer
Lead engineer on 3 production AI products: Kshetra.ai, SDR Wati, Voice SDR
- —I was embedded with founders and ops teams and owned discovery, architecture, build, deployment, and iteration on each product.
- —I was also the customer-facing engineer. I ran scoping calls, demoed weekly to stakeholders, and turned business KPIs into eval metrics.
Forward Deployed Engineer / AI Product Consultant
- —I took startup AI products from idea to production, covering voice and chat agents, LLM-powered vertical SaaS, and RAG over private data.
- —I made prototypes production-ready. That meant handling latency, cost, guardrails, observability, fallbacks, and human-in-the-loop review.
Full Stack Developer
Full Stack Developer
Earlier career
I worked as a full-stack and PHP developer for 7 companies across health-tech, travel, media, and logistics. I started freelancing in 2016 while studying at NIT Raipur.
(07) Leadership & Stack03 areas
How I lead, build, and ship AI
01 · Leadership7 skills
I turn an idea into a plan your board can back.
- A roadmap and build-vs-buy calls you can defend
- An MVP scoped to what earns revenue first
- Vendors chosen on cost and risk, not hype
- Clear updates for investors and the board
- 04 ScaleTeam, process and investor updates
- 03 LaunchLive with paying users
- 02 MVPThe one workflow that earns revenue
- 01 IdeaWhat to build, and what not to
Technical roadmappingArchitecture and build-vs-buy decisionsMVP scopingVendor evaluationCode standards and CI/CDInvestor and stakeholder communicationHuman-in-the-loop AI design
02 · AI6 skills
If you want it built with AI, I'm your man.
- 3 AI products shipped to production: Kshetra.ai, AI SDR, AI Receptionist
- The right model for each job: cost, speed and accuracy
- Every answer checked before a customer sees it
- Hands off to a human when it should
- SalesQualifies every lead
- SupportAnswers 24/7
- VoiceTakes your calls
- DocumentsFinds answers in your files
- OperationsRuns manual work overnight
- ProductAI inside your SaaS
OpenAI, Claude, GeminiLangChain, Hugging FaceRAG, evals, and prompt versioningVapi and Twilio for voiceSpeech-to-text and text-to-speech pipelinesWhatsApp Business API
03 · Engineering4 skills
Built to last after I hand it over.
- Solid data and backend foundations
- A product your users actually touch
- Standards and a deploy process your next hire can follow
Python, Node.jsNuxt / Vue, LaravelMongoDB, Postgresn8n
(08) ContactFractional CTO · AI MVP Sprint · Tech Audit
Have an idea, a prototype, or a codebase that needs a CTO?
Message on LinkedIn↗
lakshya@legit.globalLinkedInlinkedin.com/in/lakshya-verma
Instagraminstagram.com/lakshya.cto
Studiolegit.global
Based inDelhi NCR, India
Lakshya VermaFractional CTO for Startups
000Delhi NCR, India
28.61° N, 77.21° E
28.61° N, 77.21° E