NEW
Introducing Advanced AI Proctoring - The Future of Hiring Learn More

How to Hire AI/ML Talent Consistently as a Startup | Neha Kothari

Nikita Saini Nikita Saini, Author

Featured Image
GUEST PROFILE
Neha Kothari, HR Manager at Applicate AI (SalesCode.ai)
Connect: LinkedIn

This episode was hosted by Amrit Acharya, Co-Founder and COO of Xobin.

TL;DR – Key Takeaways

  • 82% of Indian employers report difficulty filling roles in 2026, with AI skills now the hardest to source globally for the first time (ManpowerGroup Talent Shortage Survey, 2026). Consistent process is the only sustainable answer to a scarce market.
  • A five-step hiring process for tech roles is not excessive. For roles like data science and AI/ML engineering, each filter serves a real purpose: skills assessment, complex coding, technical interview, cultural fit and HR round.
  • Protecting interviewer time is a hiring manager’s most undervalued job. Pre-screening by HR before candidates reach technical panels reduces hiring manager fatigue and keeps conversion rates high.
  • Consistency matters more than brand in startup hiring. Candidates judge company culture by how transparent and predictable the process is before they see the offer.
  • Upskilling that comes with genuine autonomy and senior support retains early-career tech talent more effectively than salary increments alone.

Hiring AI and ML engineers is hard for everyone. Hiring them consistently at a startup with no established brand across multiple niche roles like data science, full-stack and AI/ML is a different challenge entirely.

SalesCode.ai, an AI product firm that builds intelligent assistants and bots for FMCG and CPG companies, now rebranded as SalesCode.ai, did exactly that. When Neha Kothari joined, there was no HR department and no hiring process. Four years later, the company had scaled from 35 to over 100 employees, with a consistent pipeline of technical talent in some of the market’s most contested roles.

In this episode #4 of Xobin Talks, host Amrit Acharya, Co-Founder and COO of Xobin, sat down with Neha. They explored the mechanics behind that consistency. How the process works. How interviewer fatigue is managed. What candidates actually want. And how a startup with no brand equity builds one through its hiring experience.

How Do You Build a Consistent Hiring Process for Niche Tech Roles?

Consistency in process is the only sustainable competitive advantage a startup has in a scarce talent market. The numbers are stark. 82% of Indian employers report difficulty filling roles in 2026. AI skills now overtake traditional engineering as the hardest to source globally (ManpowerGroup Talent Shortage Survey, 2026). A startup without a structured process in that environment is fishing without a net.

Neha’s hiring architecture for technical roles runs to five steps, and she is clear that each step earns its place.

  • Step one is a skills assessment through an external platform. For SalesCode.ai, this has been Xobin from early on, enabling objective scoring before any human interviewer time is committed. 
  • Step two is a more complex coding assessment that tests applied problem-solving rather than foundational knowledge. 
  • Step three is the technical interview with the relevant hiring manager. 
  • Step four is an HR interview focused on cultural and behavioral fit. 
  • Step five is a final decision round.

The funnel this produces is significant. 200 candidates entering at assessment may produce just 10 who reach the HR round. That 5% conversion is not a sign of overly harsh filtering. It is the intended result of a process designed to protect every person’s time, including the candidate’s.

For lateral and mid-level hires, the process compresses. The assessment step reduces or consolidates, since the technical interview becomes the primary evaluation mechanism. For senior and managerial roles, the assessment is removed entirely. It is replaced with deeper technical and leadership interviews. These reflect the judgment and accountability demands of the role.

Neha built this entire process from scratch at SalesCode.ai, with no prior HR infrastructure, which gives her a practitioner’s perspective on what actually works versus what looks good in an HR textbook.

How Do You Prevent Interviewer Fatigue in a Multi-Step Technical Hiring Process?

Protecting hiring manager time is not an administrative nicety. It is a strategic decision. It determines whether your technical process delivers consistent signals or rushed, noisy assessments. In technical hiring, 50% of managers say interviewing is the toughest part of the process (Insight Global, 2025). The cognitive load compounds when the same people run multiple rounds across multiple open roles.

Neha’s solution to this is upstream, not downstream. The HR team carries more of the filtering burden before any candidate reaches a technical interviewer.

The mechanism is a pre-screening call by HR that functions as an informal first round. Before shortlisting anyone for a technical panel, Neha’s team evaluates last-minute disqualifiers: red flags that would eliminate an otherwise strong-on-paper candidate regardless of their technical performance. Salary expectations out of range, notice periods incompatible with urgency, or history that raises questions the technical panel cannot explore. Catching these before the technical round saves the hiring manager from spending an hour on a candidate who was never truly viable.

The second mechanism is careful hiring manager matching. For every open role, the HR team has an internal discussion about who the right interviewer is. This is not always the most senior person available. It is the person whose expertise maps most directly to what the role actually requires. A data science role does not need the CTO in the room if there is a principal data scientist better positioned to evaluate the candidate’s depth.

Hiring managers then spend their time on candidates who have already cleared a genuine bar. Not on those who look promising on paper but unravel during basic screening.

How Has Virtual Hiring Changed the Tech Recruitment Process?

Virtual hiring did not just move the interview online. It changed who you can reach, who shows up, and how you read a candidate. Startup hiring in India surged 32% year-on-year in April 2025. IT services accounted for 32% of all job postings (India Employer Forum, 2025). Much of that growth was enabled by the geographic expansion virtual hiring made possible.

Neha frames the shift with genuine nuance. She does not default to “virtual is better” or “in-person was better.”

Face-to-face campus drives had one structural advantage. Only serious candidates made the trip. The time investment of physically attending a drive filtered out people who were casually interested. Conversion rates from shortlist to offer acceptance were higher because attendance itself was a signal of intent.

Virtual hiring reversed this. Candidates can now be assessed from anywhere. This expanded the talent pool significantly. For niche roles, the right person might be in Pune while the team is in Delhi. But it also removed the friction filter. A candidate can attend a virtual assessment while simultaneously attending another company’s process in a different browser tab.

Neha’s response to this was to mandate video calls at every stage, not just the final round. A camera-off candidate in an early screening is a signal she takes seriously. Seeing expression, engagement and body language during a technical discussion gives HR information that audio-only cannot. It also begins building the relationship that makes a candidate more likely to accept an offer.

For candidates located in the same city as the SalesCode.ai office and who have already cleared all virtual rounds, a final in-person meeting is preferred where practical. Not a full-day experience. A direct conversation. Because some things that cannot be fully captured virtually are captured quickly and reliably when two people are in the same room.

Which Colleges Have Produced the Best Technical Hires for a Product Startup?

The most useful college list is not a prestige ranking. It is a performance-tracked record of which institutions have produced candidates who actually delivered in the role. Only 46% of graduates tested are considered employable in AI/ML roles in India (Abekus.ai, 2025). That means the right institutional targeting matters far more than broad campus coverage.

Neha is direct that the answer depends on what a company is hiring for and how much volume it needs.

For engineering and technical roles at SalesCode.ai, Grade A institutions, specifically IITs, NITs and select Delhi University programs, have produced strong candidates for specialized data science and AI/ML positions. But Grade B institutions have actually produced higher volumes of successful hires because the absolute number of candidates there is larger and competition for them from large IT firms is less intense.

Within the mid-tier college segment, IP University’s affiliated colleges, Bennett University and ABES Engineering College, have produced reliable technical hires for SalesCode.ai specifically. These are not universal endorsements. They reflect SalesCode.ai’s particular experience, its hiring criteria and the specific roles it was filling. Any company building its own campus list should run the same exercise: track cohort performance over 6-12 months per institution and let the data build the list, not third-party reputation.

For MBA-level hiring, experience at SalesCode.ai covers institutions ranging from Grade A schools like FMS Delhi and the Indian Institute of Foreign Trade to mid-range programs at IBS and similar colleges. The lesson in both cases is the same: institutional tier is a starting hypothesis, not a conclusion. Performance data is the conclusion.

How Do You Retain Technical Talent at a Startup Without a Big Brand?

Retention at a startup depends on two things that a brand cannot replace: genuine learning opportunity and the visible support of senior leadership. Startup hiring in India is increasingly skill-focused, but retaining those skills requires a different offer than simply paying market rate (ManpowerGroup, 2026). Salary is table stakes. What keeps early-career technical talent at a startup is the sense that they are growing faster than they would anywhere else.

Neha’s retention philosophy at SalesCode.ai is built around demonstrating growth velocity, not promising it.

The first element is role fluidity 

In a startup of 35-100 people, someone who is excellent at their primary function can propose moving into an adjacent area, demonstrate capability there, and make the move with the company’s support. That kind of internal mobility exists in large organizations on paper. In a startup it actually happens, and early-career professionals who experience it understand quickly that they are accumulating skills and experience at a rate that a structured corporate role would not allow.

The second element is senior accessibility

When a data scientist can discuss a model’s performance directly with the CTO, or an engineer can raise a technical concern with the founding team, the learning that results is qualitatively different from what happens in a tiered structure. That accessibility is a genuine retention lever, not just a culture statement.

The third element is market awareness 

Compensation must track the market, even if it cannot always lead it. An employee who knows they are being paid fairly relative to their skills and stage, and who is growing faster than market peers, has two good reasons to stay. Remove either one and you have a candidate who is quietly updating their resume.

What Do Tech Candidates in Their 20s Actually Prioritize in a Job?

Transparency in the process is the strongest signal a company sends about its culture before a candidate ever walks through the door. In a competitive market where candidates are evaluating multiple opportunities simultaneously, how a company runs its hiring process is how candidates judge how the company runs everything else.

Neha identifies three things that consistently matter to candidates, especially those in the 20-25 age range who are navigating their first or second significant career decision.

The first is process transparency

Candidates want to know what step they are in, what comes next, what the timeline looks like, and why they are being asked to do each thing. A company that runs a five-step process and explains each step builds more candidate confidence than one that runs a three-step process and leaves candidates wondering what is happening. Clarity about the process is interpreted as a signal about clarity in the workplace.

The second is pre-interview preparation support 

Candidates who receive meaningful company information before they arrive at an interview, not just a job description but context about the product, the team, the business and what success in the role looks like, perform better and feel more respected. A candidate who can engage with the interviewer as an informed peer, rather than as someone hearing about the company for the first time, is a better interview for everyone. Providing that information proactively is a small investment with a significant return on candidate experience.

The third is consistency

Market fluctuations create pressure on companies to adjust their offer standards, their role definitions and their process timelines rapidly. Candidates experience this inconsistency and draw conclusions from it. Neha’s position is that a company should be clear about what it actually needs, hire for that, and maintain those standards regardless of what the market is doing around it.

What Is the Right Recruitment Tech Stack for a Growing Startup?

Every tool in your recruitment stack should be justified by its measurable output. Not by the fact that a larger company uses it. 83% of companies plan to use AI for resume screening by 2025. But adoption without ROI discipline creates cost without clarity (Resume Builder, 2025). For a small HR team, tool selection is a resource decision as much as a technology one.

Neha’s stack at SalesCode.ai covers three functional categories.

The first is a skills assessment platform 

For technical roles, an objective skills assessment before any human interviewer is involved is non-negotiable. It removes resume-based bias from early shortlisting and ensures that candidates who reach the technical round have cleared a baseline that is relevant to the role, not just to their academic credentials.

The second is sourcing platforms

Knowing where your specific candidate profiles are most reachable, whether LinkedIn, specialized job boards or college placement portals, matters more than having accounts on every platform available.

The third is an ATS

Even for a company of 50-100 people, tracking candidates across a five-step process without a centralized system creates coordination failures that cost time and candidate goodwill.

Neha’s principle for justifying tools to management is direct: when requesting renewal or expansion of a recruitment tool, come with data showing what it produced. Which assessment platform identified candidates who went on to perform well? Which sourcing channel produced the highest quality-to-volume ratio? ROI is not optional when asking a startup leadership team to fund HR infrastructure.

🎧 Watch the Full Episode

Xobin Talks – Episode 4 | Neha Kothari, HR Manager, SalesCode.ai | Hosted by Amrit Acharya, Co-Founder and COO, Xobin

▶ Play Episode #4 of Xobin Talks

Did you enjoy this episode of Xobin Talks? Like, share, and join the Xobin HR community for more insights from leaders.

About Neha Kothari

Neha Kothari is the HR Manager at Applicate AI, now rebranded as SalesCode.ai, a fast-growing AI product company building intelligent assistants and bots for FMCG and CPG enterprises. She built the HR function at SalesCode.ai from the ground up, having joined when there was no HR department at all. 

Over four years, she scaled the company’s operations from a 35-person team to over 100 employees, hiring consistently across demanding technical roles, including data science, AI/ML engineering and full-stack development. Her background spans HRIS consulting at Infosys, HR in FMCG with blue-collar workforce management, and now product-first tech startup HR. She holds an MBA from BHU (Banaras Hindu University).

Connect with Neha on LinkedIn | Company: SalesCode.ai

Want more insights like this? Explore all Xobin Talks episodes.

Frequently Asked Questions

  1. How many interview rounds are appropriate for tech hiring at a startup? 

    Five steps for specialist roles: skills assessment, complex coding test, technical interview, HR interview and final decision. Each step serves a specific filter purpose. For lateral or senior hires, the assessment step reduces or is removed and replaced with deeper technical interviews.

  2. How do you prevent interviewer fatigue in a multi-step hiring process? 

    HR pre-screening eliminates last-minute disqualifiers before any candidate reaches the technical panel. Salary expectations, notice periods, communication gaps and basic red flags are caught upstream. Hiring managers then only interview candidates who have already cleared a genuine bar.

  3. What changed most in tech hiring after the pandemic? 

    The talent pool expanded geographically, which is a net positive for niche roles. But the friction filter disappeared. Candidates now attend multiple processes simultaneously. Mandatory video calls at every stage and a final in-person meeting for local candidates partially offset this loss of commitment signal.

  4. Which colleges produce the best technical hires for a startup? 

    Grade A institutions produce strong specialists. Grade B colleges produce higher volume at lower competition from large IT firms. Track cohort performance per institution over 6-12 months and build your list from that data, not from external reputation rankings.

  5. What do young tech candidates in their 20s prioritize in a job? 

    Three things: a transparent hiring process that signals what workplace clarity will feel like, pre-interview company context so they arrive prepared rather than surprised, and consistency in standards regardless of market fluctuations. Money matters, but these signals shape trust before an offer is even made.

  6. How do you retain early-career tech talent at a startup without a big brand? 

    Through demonstrable growth velocity: role fluidity that allows movement into adjacent areas, direct access to senior leadership for learning, and compensation that tracks the market even if it cannot lead it. Freshers who grow faster than peers have two reasons to stay. Remove either one and you have a flight risk.

  7. What tools should a startup HR team invest in for recruitment? 

    Three categories cover the essentials: a skills assessment platform for objective pre-screening, sourcing platforms matched to your candidate profiles, and an ATS for tracking candidates across multiple steps. Justify every tool renewal with measurable output data, not feature comparisons.

  8. What is Neha’s view on moonlighting for tech professionals? 

    It depends entirely on the employment agreement. If the agreement permits it and the company accepts it, that is the employee’s right. If it does not, proceeding without transparency is the problem, not the side work itself. The risk is divided attention and conflicting responsibilities, which ultimately harms both parties.

Leave a Comment

Nikita Saini

Nikita Saini

About the author

Nikita writes practical and research-based content on Psychometric Testing, Interviewing Strategies, and Reviews. Her work empowers hiring professionals to enhance candidate evaluation with a structured, data-informed approach.

Discover the Power of Efficient Candidate Assessments

Get started with Xobin today, streamline your hiring process and hire your ideal candidates.

Get Started
Marketing CTA