Hi, I'm Nijil (/ˈnɪdʒɪl/), you can call me Nij.
I'm a GTM leader that helps high-growth companies find their market, launch new products, and scale their sales motion.
At Merge, I own product marketing across the Merge for Product suite (unified API, MCP gateway, and LLM routing).
Before that, I shaped GTM at Sigma Computing for AI, apps, and analytics products.
Positioned a model router as enterprise infrastructure instead of leading with cost-savings. The follow-up launch for routing customizability saw the founder’s launch post reach 3.6M views.
Introducing Merge Gateway - Build Your Own Router.
You're three sprints into your coding assistant.
You pick the most hyped model, integrate, test, deploy.
A month later, a new model drops.
Now you re-test, re-integrate, re-deploy.
Your product didn't change, but the… Show more
Every competing proxy opened with savings. Uptime and enforced budgets answer the two things that actually stall adoption: a provider going down, and a bill nobody can predict.
Engineering leads will not route production traffic through a black box. Customer-set benchmark weights, DLP rules and a visible request log make the control configurable and auditable rather than asserted.
Per-customer routing, BYOK and pass-through billing are what a platform needs to resell routing to its own customers. No cost-saving proxy ships this, and it moves Gateway out of that comparison entirely.
Model routers help products access the latest models without updating code, avoid relying on one provider, and optimize spend. Competitors were leading with cost reduction, targeting the developer audience, and requiring their customers to trust them with routing decisions.
Merge led with enterprise control, reliability, and routing customizability. Build Your Own Router let customers set their own benchmark weights, upload internal evals, and keep an audit trail.
Hundreds of thousands in qualified pipeline traced to the launch page, and dozens of enterprise teams deployed custom router configurations in the first 30 days.
Merge was known as a unified API before building its Gateway and Agent Handler products. I turned these three core capabilities into a two-product story that made up the connectivity layer for enterprise AI.
Developers shipping AI features and IT leaders governing AI access land on the same page. Naming the layer instead of the product let both see themselves in it.
Forking into two products at the top would have read as two companies. One claim first, then the pillars, so Merge stays one company with two doors.
Agent projects die in security review, so scoped permissions and audit logs sit in the first thing a developer reads instead of buried on a trust page.
Competitors sold cost reduction. Leading with failover and customer-set rules answers the real fear, which is being locked to one provider.
Unified API had to read as a pillar of the AI story rather than the legacy business. It sits third on purpose: earned, not headlined.
Permissions ride on Okta and Entra rather than a new system to govern. The pitch is less work, not another console.
Access alone leaves employees rebuilding work that already exists. A company-wide context layer gave IT a reason to buy beyond permissions.
Governance language does not move IT. Per-team spend caps and an approved model list turn policy into something a buyer can picture enforcing.
Security signs off last and can kill the deal, so DLP and redaction end the sequence rather than open it.
Merge was known as a unified API. But customers building agents needed MCP connectors, scoped authentication, and LLM routing, not another data sync.
Two pillars, so the buyer journey never forked mid-page. Merge for Products, for teams shipping AI applications. Merge for Workforce, for IT/security and AI leaders connecting internal tools to enterprise systems securely.
Millions in open qualified pipeline traced to the new narrative, across IT, SecOps and AI transformation.
Repositioned a cloud analytics provider as operational software, and built the first ever virtual conference around the category. Became the highest lead-generating marketing event in company history.
Naming Data Apps in a blog post makes it a claim. Building a conference around it makes it a category, with a definition on the page so the term could not drift.
Blackstone, Conagra, WHOOP, and DoorDash describing their own apps builds credibility. The session wall was the proof layer of the launch.
Blackstone, Conagra, WHOOP, and DoorDash describing their own apps builds credibility. The session wall was the proof layer of the launch.
Blackstone, Conagra, WHOOP, and DoorDash describing their own apps builds credibility. The session wall was the proof layer of the launch.
I designed and produced each partner video so integrators could showcase their own builds. That deepened the partnerships and pushed Data Apps into industries we could not staff alone.
Every BI tool sells the same dead end. Read a chart, spot a problem, then open five other systems to do anything about it.
Sigma's direct warehouse writeback broke that, so we named what it made possible. Data Apps (build, automate, act) put Sigma in the operational software conversation for RevOps, FP&A and logistics rather than the analytics one.
Hit 118% of the target audience goal, hundreds of SQLs from the inaugural conference, and dozens of templates in customers' hands.
Developers judge infrastructure by the integration, so the code block leads. Both CTAs, Start building for free and Read the docs, point at the product and the docs rather than a sales form.