How to Prep for Product, Service, Analytics, and AI-First Companies

Most engineering students use the same preparation playbook for every company on their placement list. That is where the effort leaks.

The Four Company Buckets

Placement drives in India divide into four distinct hiring models. Each tests a different skill mix, sets different cutoffs, and rewards a different preparation strategy. Getting the taxonomy right before you start saves weeks of misdirected effort.

Bucket	Representative companies	Primary test weight
Product	Google, Microsoft, Amazon, Zoho, Freshworks	DSA and system design
Service	TCS, Infosys, Wipro, Cognizant, HCL	Aptitude and basic tech
Analytics	Mu Sigma, ZS Associates, LatentView, Fractal Analytics, Tiger Analytics	Quant reasoning and case analysis
AI-first	TCS Prime track, Quantiphi, SigTuple, and AI engineering roles at mid-size product companies	Deployed projects and Python depth

The four-bucket frame does not mean you prep in total isolation for each one. There is a shared base (aptitude concepts, one programming language, basic data structures, and communication skills) that all four demand. The specialisation happens in the final 10 to 12 weeks before your placement window.

Product Companies: DSA-First Hiring

Product companies, meaning firms whose revenue comes primarily from software they build and own rather than services they bill to clients, weight algorithmic problem-solving above almost everything else at the interview stage.

Google, Microsoft, and Amazon run 4 to 6 rounds of technical interviews for fresh hires, focused on data structures, algorithms, and basic system design. Domestic product companies like Zoho and Freshworks are lighter on distributed system design at the fresher level but still expect solid DSA fluency.

What to build during prep:

• Arrays, linked lists, trees, graphs, heaps, and hash maps (data structures)
• Sorting, searching, dynamic programming, greedy, and graph traversal (algorithms)
• Object-oriented design basics, particularly for Zoho and Freshworks rounds
• Aptitude as a pass/fail filter; not a ranking mechanism

The scoring dynamic is different from service companies. Clearing aptitude gets you into the interview pipeline at product companies. DSA rounds are where candidates are actually ranked. Solve problems at LeetCode medium difficulty consistently and you’re competitive for domestic product companies; the bar rises sharply for FAANG-tier roles.

Service Companies: Volume Hiring on Aptitude

Service companies screen at scale using standardised tests. TCS runs the NQT (National Qualifier Test), with your score placing you into one of three tracks:

• Ninja: ₹3.5 to 3.9 LPA
• Digital: ₹7.0 to 7.5 LPA
• Prime: ₹9.0 to 11.0 LPA

Understanding the TCS NQT section pattern before starting prep cuts wasted study time.

The NQT covers verbal ability, reasoning, numerical ability, programming logic, and a coding section. For Ninja, the aptitude sections carry the most weight. For Digital and Prime, coding ability and AI skills are increasingly the differentiators.

TCS CHRO Sudeep Kunnumal stated at the AI Impact Summit in March 2026 that 60% of TCS’s FY26 fresher hires are AI-skilled, up from 10 to 15% three years ago. Total FY27 fresher intake is projected to drop to around 25,000 from 44,000 onboarded in FY26, per Financial Express, with a heavier tilt toward AI-capable candidates. The Ninja track gets more competitive as total volume contracts.

For Infosys, Wipro, and Cognizant, the pattern is similar: an online aptitude test (InfyTQ for Infosys, Wipro NLTH, or eLitmus-driven screening) followed by technical and HR rounds. Spending four to six weeks drilling quantitative aptitude shortcuts before placement season compounds across every service company drive simultaneously.

The TCS coding section tests two problems in 45 minutes. Array and string manipulation at Ninja level; more complex DP or graph problems at Digital and Prime. Thirty to forty problems at consistent medium difficulty covers the coding gate for most service company drives.

Analytics Companies: Quant Depth Over Breadth

Analytics companies are often miscategorised as light-touch service hiring. The process is nearly the inverse of service company hiring: less emphasis on aptitude speed, much more on quant reasoning depth and structured problem decomposition.

Major analytics employers from the India campus circuit:

• Mu Sigma — quant-heavy written test, case rounds testing structured problem breakdown
• ZS Associates — case interviews, logical reasoning, and business analytics scenarios
• LatentView Analytics — SQL, basic Python, statistical reasoning
• Fractal Analytics — data interpretation, basic ML concepts, problem-structuring cases
• Tiger Analytics — Python or R basics, analytical writing in assessment rounds

What to build during prep:

• Quant reasoning at a level harder than standard placement aptitude (data sufficiency, multi-condition probability, complex permutation scenarios)
• SQL: joins, aggregations, window functions, not just SELECT basics
• Python basics for data manipulation (pandas, numpy), not competitive algorithms
• Case analysis: decomposing a business problem into measurable components with a logical chain

Students preparing only for analytics companies can deprioritise classical DSA nearly entirely. The tradeoff is that this prep does not transfer well to product-company interview rounds.

AI-First Companies: The Emerging Fourth Track

Since roughly 2022, a distinct fourth hiring model has taken shape in India: companies and roles that specifically target candidates who can build with AI tools, not just describe them on a resume.

This includes TCS Prime, which now reviews AI or data project portfolios at the ₹9 to 11 LPA Prime tier (per TCS CHRO Kunnumal, March 2026), along with Quantiphi, SigTuple, and AI-focused engineering roles at mid-size product companies that did not exist five years ago.

What distinguishes AI-first hiring:

• A deployed or publicly accessible project (GitHub repo, Colab notebook, small API) carries more weight than a certification
• Python fluency, not deep algorithm mastery, but confident with pandas, APIs, and basic LLM wrappers
• Awareness of current tooling: LLM APIs, vector databases, basic model evaluation
• Aptitude and communication gates still apply; they are not waived for AI-track candidates

This track does not replace the other three. It overlaps with product companies (if the firm builds AI products) and with analytics (for data-science-adjacent roles). The clearest signal that a role is AI-first is a JD that asks for “project experience” rather than just “internship experience.”

Routing Your Prep: A Time Allocation Framework

Here is a practical allocation across a 6-month runway before placement season, by target bucket.

Phase	Shared base (all tracks)	Product specialist	Service track	Analytics track	AI-first track
Month 1 to 2	Aptitude concepts plus basic DSA	—	—	—	—
Month 3 to 4	Aptitude practice plus one language fluent	LeetCode 75 problems	NQT mock tests	SQL and case practice	Build one project
Month 5	Communication prep	DSA medium to hard depth	Company-specific patterns	Quant depth drills	Iterate and document project
Month 6	Full-length mock tests	Mock technical interviews	HR round prep	Case interview practice	Portfolio presentation practice

Students targeting multiple buckets (a service company as a floor with a product company as a stretch) should not specialise before Month 3. The aptitude base is shared and compounds with consistent practice; splitting attention too early produces two shallow preparations instead of one solid one.

A comparison of placement prep platforms maps which platform fits each prep phase and bucket best.

Standard placement prep materials don’t cover the specialisation that separates candidates at analytics and AI-first firms. TinkerLLM’s project modules start at ₹499 on tinkerllm.com and build the deployed portfolio that AI-first JDs specifically ask for, which carries more weight at TCS Prime or Quantiphi than any certification stack.