“Cracking aggressive exams like Joint Entrance Examination and Birla Institute of Know-how and Science Admission Take a look at opened doorways to BITS Pilani, certainly one of India’s premier engineering institutes,” recollects Kshitiz Parashar, Founding AI Engineer at Alhena, a SaaS startup remodeling how ecommerce manufacturers have interaction with their prospects.
“Dorm rooms doubled as incubators and that vibe made beginning an organization really feel regular,” he says, reflecting on his early years as a child who beloved ideas in Physics and Maths. Whilst scholar loans loomed, he knew a 9-to-5 job was a simply pit cease, not the vacation spot.
In an more and more saturated panorama of AI-powered customer support, most options nonetheless battle to maneuver past flashy demos and primary chatbots. However at Alhena, a breakthrough in AI engineering is reshaping what’s potential for ecommerce assist. On the coronary heart of this transformation is Parashar, whose improvements have helped propel the corporate from zero to seven figures in annual recurring income (ARR).
Parashar has constructed Alhena’s core AI agent framework from the ground-up, an clever system able to dealing with real-world buyer queries with minimal hallucination, deep model coaching, and enterprise-grade reliability. The end result: for some ecommerce manufacturers, Alhena’s brokers now resolve over 80% of incoming assist queries, outperforming legacy chatbot platforms by as a lot as 10x in accuracy, effectivity, and gross sales impression.
Engineering past prototypes
“Everyone seems to be constructing AI brokers, however only a few are placing them into actual manufacturing environments at scale,” says Parashar. “What we’ve constructed isn’t only a prototype however a production-grade, multi-agent system that powers high-reliability assist for ecommerce manufacturers.”
Whereas he’s fast to credit score his co-founders, Nagendra and Ashu for his or her imaginative and prescient and mentorship, it’s clear Parashar’s technical management has been instrumental in turning Alhena’s imaginative and prescient into actuality.
At a time when most corporations are nonetheless experimenting with retrieval-augmented technology (RAG) and scripted bots, Parashar’s contributions have made Alhena leapfrog the area via agent deployment that deeply perceive model context, reply intelligently, and scale with enterprise development. This success is rooted in improvements reminiscent of:
A hallucination detection engine that reduces misinformation in responses.An AI agentic structure designed to collaborate, escalate, and be taught constantly.A sophisticated information ingestion pipeline that auto-updates from product catalogs and buyer suggestions.
These techniques don’t simply automate replies but additionally assist manufacturers get well deserted carts, scale back churn, and switch customer support right into a income driver.
Classes of grit, failure, and studying
Parashar’s skilled journey started as a machine studying engineer in a significant financial institution. “The work was intellectually fascinating, nevertheless it lacked the real-world impression I craved for,” he says. So he teamed up with a co-founder to construct good visitors lights to deal with congestion in Indian cities, till COVID-19 disrupted all the things.
After an early stint constructing good metropolis tech in India and a Grasp’s diploma from College of California, Davis, he was a founding engineer at a voice assistant startup that was too early for the market. However the classes have been invaluable. “We failed quick, realized sooner,” he displays.
When the chance to hitch Alhena AI got here up, it felt like a pure subsequent step for the younger founder. The mission was clear: assist e-commerce companies deploy dependable, production-ready AI brokers. “Taking AI from prototypes to manufacturing has been probably the most fulfilling experiences of my profession,” he shares.
Constructing with values
Ask Parashar what drives his work, and the reply is straightforward: possession. “If you wish to construct one thing nice, it’s important to personal it finish to finish – from the preliminary thought via execution and past,” he shares. He’s additionally fast to emphasise that failure isn’t the top; it’s a redirection. “Onerous issues result in a better life. Straightforward selections normally result in more durable occasions,” he states.
At its core, Alhena AI solves a painful and costly drawback in ecommerce: poor customer support. Manufacturers lose gross sales and burn out assist groups whereas consumers abandon carts in frustration. Alhena flips this narrative by giving each service provider an AI agent that’s quick, dependable, and genuinely useful; deflecting round 86% of tickets and recovering misplaced gross sales.
Imaginative and prescient and impression
Alhena’s development is constructed not simply on innovation but additionally on values. Each buyer interplay goes via rigorous checks to make sure AI brokers behave ethically, constantly, and safely. “Our brokers are brand-aligned and human-aligned,” Parashar emphasizes. “That’s non-negotiable.”
The roadmap forward is bold: go from seven to 9 figures in ARR by changing into the de facto AI layer for ecommerce assist worldwide. This implies fixed experimentation, fast iteration, and staying forward of the most recent in LLMs, fine-tuning strategies, and agent architectures.
“Startups are about tempo. In case you’re not transferring quick, you’re already behind,” he says.
Recommendation for aspiring AI builders
Requested what he would suggest to a 22-year-old trying to break into AI, Parashar is candid. “Be a part of a startup the place you may take actual possession. Be taught quick. Remedy precise issues. You’ll develop sooner than you ever imagined,” he says.
With AI maturing quickly and expectations rising even sooner, Alhena’s story is a blueprint for the way robust engineering, clear product focus, and relentless execution can flip cutting-edge analysis into transformative impression; and for Parashar, it’s just the start.
