Anup Engineering Software Engineer Interview Guide

The 60-second brief

Candidates describe Anup Engineering as an industrial manufacturer whose work involves mechanical systems, static equipment design and large infrastructure projects. The Software Engineer role is described as sitting between software and mechanical engineering: building tools that speed up estimation, supporting electrical maintenance, and keeping technical documentation in line with safety and quality requirements. Day to day, candidates report working alongside teams for static equipment design, electrical maintenance and safety engineering. The software exists to serve physical equipment, so the people hiring for it ask about the physics and materials behind that equipment.

This guide is for candidates applying to the Software Engineer opening at Anup Engineering. It covers the three reported stages (application, initial screening, technical personal interview), the technical-fundamentals questions candidates report (iron-carbon diagram, stress-strain curve, mass density, material properties versus structural integrity, static equipment safety), and the commitment questions (the 3-year bond, handling pressure, why this company, clarifying employment terms). No coding or system design questions are reported for this loop, so the plan spends its time on engineering fundamentals and on your spoken answers.

Candidates report three stages: an application (often via employee referral), an initial screening with an aptitude test and a technical examination that carries a 40% cutoff, and a technical personal interview on applying engineering concepts live.

Focus areas

• Iron-Carbon Equilibrium Diagram
• Stress-Strain Curves
• Static Equipment Design

Overview

The reported questions for this loop are unlike a typical software loop. Five of the nine are mechanical and materials fundamentals: the iron-carbon equilibrium diagram, the stress-strain curve, how to calculate mass density, how material properties relate to structural integrity, and the safety considerations around static equipment. The other four are about commitment: willingness to accept a 3-year bond, handling high-pressure manufacturing environments, why you want this company in particular, and how you go about clarifying employment terms. Candidates also list engineering estimation and EHS (environmental, health and safety) as topics, and describe example scenarios such as analysing a material failing under stress and estimating capacity for a new project.

That mix shapes preparation. The reported responsibilities (tools for estimation, electrical maintenance support, documentation that meets safety and quality requirements) mean you should be able to explain a concept from first principles and then say what it changes in a piece of equipment or a workflow. A memorised definition of the stress-strain curve is a weaker answer than one that names yield strength, ultimate tensile strength and ductility and says which of them a designer would use to set an allowable stress.

The screening stage is the first hard gate. Candidates report an aptitude test plus a technical examination with a 40% cutoff on the technical portion. A cutoff of that kind rewards breadth over depth: you need every core topic to be solid before you spend time on rare corner cases. Check units and orders of magnitude, and revisit the basics such as density, stress and strain, and the iron-carbon phases until you can reproduce them from memory.

Finally, plan for the commitment questions as seriously as the technical ones. The bond question has a real decision behind it, so settle your own answer before the conversation, and read the terms before accepting any offer. The advice candidates pass along is to ask for the details early, including bond requirements, so that your yes or no is an informed one.

The Anup Engineering interview process

Candidates report three stages over roughly three to five weeks, though other reports describe the process finishing sooner. Treat the order below as what candidates describe, not a published process.

Application Submission

Candidates report that applications are submitted directly, and that many arrive through an employee referral. There is no technical content reported for this stage, so the work is to make the application read as a fit for a role that sits between software and mechanical or electrical engineering. If you have a contact at the company, candidates suggest using a referral.

What to demonstrate

• How your background maps to a software role supporting mechanical and electrical engineering teams
• Whether your application shows any link to industrial or manufacturing work

How to prepare

• Rewrite your CV so any mechanical, electrical, lab or plant exposure sits near the top, next to your software projects
• Write two sentences on why a software role in heavy industrial manufacturing suits you, which you will reuse for the question on why Anup Engineering
• If you know a current employee, ask for a referral and ask what the team builds before you submit

Initial Screening

Candidates describe a screening that combines an aptitude test with a technical examination, and report a 40% cutoff on the technical portion. The topics candidates list are the iron-carbon equilibrium diagram, stress-strain curves, static equipment design, engineering estimation and EHS. The contents of the aptitude section are not detailed in reports, so general quantitative practice is the safe preparation. Because a score threshold decides who moves on, make sure no core topic is blank.

What to demonstrate

• Recall of core materials and mechanical fundamentals under exam conditions
• Quantitative aptitude, as the round's name and format imply
• Whether your technical score clears the reported 40% cutoff

How to prepare

• Build a one-page sheet for each reported topic: iron-carbon diagram, stress-strain curve, density, static equipment safety, estimation
• Do timed practice on unit conversions and quick calculations, such as mass from density times volume
• Answer the fundamentals questions aloud first, then check the answer against a textbook, and repeat on the ones you missed

Technical Personal Interview

Candidates who pass screening are described as moving to a technical personal interview that focuses on applying engineering concepts in real time. The example scenarios candidates give are analysing how a material fails under stress and estimating the capacity needed for a new project. The advice that comes with it is to lay out your reasoning in clear steps, and to explain how you would work toward an answer if you do not know it.

What to demonstrate

• Live application of materials, mechanical and safety concepts to a concrete scenario
• Clarity of step-by-step reasoning when you do not know the answer outright
• Your link between a concept and its effect on equipment or maintenance workflows

How to prepare

• Practise a failure-analysis walkthrough: loading, stress level against yield and tensile strength, material condition, then the likely failure mode
• Practise a capacity estimate out loud, stating assumptions, formula, units and a sanity check
• Rehearse your answers on the 3-year bond and on why Anup Engineering, since the reported question list includes both

Where candidates lose points

Preparing as if the loop were a coding interview

The reported questions are materials, safety and commitment topics, with no coding or system design question reported. Spend your first days drawing the iron-carbon diagram, the stress-strain curve and the density formula, rather than on algorithm problems.

Reciting a definition without saying what it means for equipment

For each fundamentals topic, finish the answer with a consequence. For the stress-strain curve, say which point a designer treats as the limit and why a safety margin sits below it. For static equipment, say which hazard each safety measure addresses.

Dropping units or skipping a sanity check in a calculation

On a density or capacity estimate, write the formula, substitute with units, and check the result against a known value, for instance that steel is roughly 7,850 kg/m3. A technical exam with a 40% cutoff punishes careless arithmetic as much as missing knowledge.

Answering the 3-year bond question on reflex, or dodging it

Decide your answer in advance, and ask about the repayment terms, start point and written conditions. Say yes, yes with stated conditions, or not yet, and give the reason in one or two sentences.

Giving a generic reason for wanting this company, or a pressure story where safety checks were skipped

Tie your motivation to the reported work on estimation tools, maintenance and safety documentation. In the pressure story, name the check you kept even under time pressure, and say whom you told when the date moved.

Questions & practice

Diff a projection against the primary without per-row point reads

Hard · reconciliation · range hashing · throttling

The listing projection has drifted and some rows show a stale version. The primary holds 40,000,000 resource rows across 12,000 tenants while serving 1,200 writes and 14,000 reads per second. The obvious repair, reading each resource row and comparing its version against the projection, is correct and would eventually finish. Explain precisely why it is unacceptable here, then give a diff that finds the differing rows, state its complexity, and make it safe to run against a live primary. Replication lag is usually under 100 ms and is not bounded.

Approach

1. Quantify the naive cost rather than calling it slow: 40,000,000 point reads at even 0.5 ms each is over five hours serialised, and the only lever is concurrency, which is exactly what you cannot spend. The primary's pool is sized for the write path, and 40,000,000 random reads evict the buffer cache that sustains the 85 percent cache hit rate, so the audit degrades the system it is auditing.
2. Replace random access with one ordered pass per side. Both sides can be read in (tenant_id, resource_id) order, which is a sequential scan on each and a merge join in O(n) time and O(1) memory. For a dense diff that is the whole answer, and it reads the primary once instead of 40,000,000 times.
3. For the expected sparse case, compare range hashes instead of rows: partition the key space, compute per range an order-independent aggregate over hash(resource_id, version), compare aggregates, and descend only into ranges that differ. With d differing rows and branching factor B, at most d ranges mismatch per level, so the drill-down examines O(d log_B(n/d)) ranges and reads full rows only in mismatching leaves.
4. Aggregate with a sum modulo 2^64 or a multiset hash, never XOR. XOR is order-independent but self-cancelling, so two rows wrong in the same way, or a row duplicated on one side, leave the range aggregate matching and the range is declared clean.

Follow-up

• The diff reports 900 stale rows. How do you decide between patching those rows and rebuilding the projection from resource_revision?
• Same job, but the projection lives in a search index that cannot be scanned in key order. What changes?

Merge partitioned event streams into one ordered feed with bounded lateness

Hard · k-way merge · watermarks · out-of-order streams

The read-model service consumes 64 log partitions carrying about 4,000 events per second in total. Each partition is ordered within itself, but partitions drift by up to 30 seconds, and the activity feed must present a tenant's events in occurred_at order. Produce the merge. State its complexity, the buffer it requires in events and in bytes, what happens when one partition is idle, and what you do with an event that arrives after you have already emitted its position. Payloads average 1 KB.

Approach

1. Merge with a min-heap over the 64 partition heads keyed on (occurred_at, event_id): O(log P) per event and O(n log P) overall. The tie-break on event_id is what makes the output deterministic when two partitions carry the same millisecond, which matters because the feed is paginated and a non-deterministic order reorders pages under the reader.
2. Emitting the heap head is only correct once every partition has produced everything up to that timestamp, so the emit condition is a watermark: the minimum across partitions of the highest occurred_at seen, less the allowed lateness. Events are held until the watermark passes them, which is what turns individually ordered streams into a jointly ordered one.
3. Size the buffer from the lateness rather than guessing: 4,000 events per second times 30 seconds is 120,000 buffered events, and at 1 KB each about 120 MB of heap. That number is the real price of the ordering guarantee and belongs in front of whoever asked for it.
4. Handle the idle partition explicitly, because it fails the feed rather than corrupting it: a partition with no traffic never advances its own maximum, so the watermark freezes and output stops entirely. Either every partition emits a periodic idle marker carrying the broker's current time, or the watermark falls back to wall clock for a partition silent beyond a threshold.

Follow-up

• The lateness budget is raised to five minutes. What is the new buffer, and what besides memory changes?
• The consumer restarts. Where does it resume from, and what does the feed look like for the first 30 seconds?

Identify the heaviest tenants in a five-minute window under memory pressure

Medium · top-k · heavy hitters · streaming

The edge service handles about 3,000 requests per second across roughly 50,000 tenants, peaking near 9,000. Expose the 50 heaviest tenants by request count over the trailing five minutes so limits can be tightened before one tenant's backfill starves the fleet. You may not retain five minutes of raw records. Give the exact solution and its memory, then the bounded-memory approximation with its error stated as a formula, and say which you would ship and at what tenant cardinality that choice changes.

Approach

1. Do the exact version first, because it is affordable at this cardinality: a ring of 300 one-second counters per tenant, advanced lazily, is 1,200 bytes of counters per tenant and roughly 60 to 90 MB for 50,000 tenants with overhead. Carry a running total and subtract the bucket you overwrite so a window read is O(1) rather than 300 adds.
2. Extract the top 50 with a size-k min-heap over the tenant sums: O(d log k) for d tenants, against O(d log d) to sort them all. Maintaining the heap continuously instead of on query requires a tenant-to-heap-index map, because incrementing a count already inside the heap means sifting from a known position, and without that map you rebuild the heap on every request.
3. State the approximation precisely rather than gesturing at sketches. Misra-Gries with m counters retains every item whose true count exceeds N/(m+1), and each retained count underestimates the truth by at most N/(m+1). With m = 1,000 and N = 900,000 requests in the window the error is roughly 900 requests, which is fine for spotting a tenant sending 50,000 and useless for ranking two tenants 200 apart.
4. Say what breaks when the window slides: Misra-Gries and Space-Saving are insert-only and cannot be decremented as records age out. The workable construction is one summary per sub-window, say ten seconds, with 30 summaries merged at query time, and the merged error is the sum of the per-summary errors, so the bound degrades linearly in the number of sub-windows.

Follow-up

• The heaviest tenant is heavy because of one export job rather than user traffic. Should the limiter treat those as the same tenant?
• Two tenants sit tied at the boundary of the top 50. Does your answer flap, and does the flapping matter?

Find version gaps and relay lag with window functions

Medium · window functions · event tables · frame semantics

outbox_event holds event_id, aggregate_type, aggregate_id, aggregate_version, event_type, payload, status ('pending','published','dead'), attempts, created_at, published_at. A projection is missing rows and you must decide whether the relay skipped events or the consumer dropped them. Write three queries over the last seven days: one listing every aggregate_id whose published aggregate_version sequence has a hole, one giving per-day counts with a running total, and one returning the newest published event per aggregate. For each, say where the window function is evaluated relative to WHERE and LIMIT. PostgreSQL 16.

Approach

1. Gaps: compute lead(aggregate_version) OVER (PARTITION BY aggregate_id ORDER BY aggregate_version) in a subquery, then filter next_version <> aggregate_version + 1 in the outer query. Window functions are evaluated after WHERE, GROUP BY and HAVING and before the outer ORDER BY and LIMIT, so the predicate cannot sit in the same WHERE clause and PostgreSQL 16 has no QUALIFY.
2. Say what the seven-day filter does to the answer: it truncates every partition, so the first row per aggregate has no predecessor inside the window and a hole spanning the boundary is invisible. Widen the window, or join to resource.version as the authority for the true maximum.
3. Running total: SELECT date_trunc('day', created_at) AS d, count() AS n, sum(count()) OVER (ORDER BY date_trunc('day', created_at) ROWS UNBOUNDED PRECEDING). An aggregate inside a window call is legal because grouping runs before windowing. The grouping key is unique per row here so ROWS and RANGE agree, but write the frame anyway — over ungrouped rows with tied timestamps the default RANGE frame pulls in every peer row and the total jumps.
4. Newest per aggregate: DISTINCT ON (aggregate_id) ... ORDER BY aggregate_id, aggregate_version DESC is the cheap PostgreSQL-only form when an index matches that order; row_number() OVER (PARTITION BY aggregate_id ORDER BY aggregate_version DESC) = 1 is the portable form and needs a subquery for the same evaluation-order reason as the gap query.

Follow-up

• Relay failover redelivers events. Does a duplicate break the gap query, and how would you detect one from this table alone?
• Turn the gap check into a continuous monitor rather than a query someone runs after an incident. What does it watch?

Hold a per-tenant active cap against concurrent creates

Hard · write skew · isolation levels · constraints

A tenant on the standard plan may hold at most 50 resources with status='active'. The create handler runs SELECT count(*) FROM resource WHERE tenant_id = $1 AND status = 'active', compares to 50, then inserts. Two creates arrive 3 ms apart on different instances and the tenant lands at 51. Name the anomaly, say whether PostgreSQL 16 READ COMMITTED or REPEATABLE READ prevents it and why, then give an implementation that holds the cap at READ COMMITTED with the exact statements. Finally, say what changes when the cap is 'at most one running export per tenant' on job_run.

Approach

1. Name it: write skew. The two transactions read an overlapping set and write disjoint rows, so there is no row-level conflict for the engine to detect and each commit is individually legal.
2. Rule out the levels precisely. READ COMMITTED takes a fresh snapshot per statement and takes no lock on the counted rows, so both see 49. PostgreSQL's REPEATABLE READ is snapshot isolation: it removes non-repeatable reads and phantoms within the snapshot but still admits write skew, because the anomaly is not a re-read of a changed row, it is a read of a set that a concurrent transaction invalidates. Only SERIALIZABLE closes it, by tracking the read dependency and aborting one transaction with SQLSTATE 40001 — a guarantee that exists only if the application re-runs the whole transaction from the read.
3. Convert the set predicate into a single-row conflict: keep tenant.active_resource_count and run UPDATE tenant SET active_resource_count = active_resource_count + 1 WHERE tenant_id = $1 AND active_resource_count < 50 in the same transaction as the INSERT. Zero affected rows is the cap, returned as 409. The row lock serialises the decision at any isolation level, and contention is bounded to one tenant's row — which is also the fair-scheduling unit, unlike a global counter that would convoy every tenant behind one row.
4. State the cost you just took on: a counter is a second source of truth that can drift, so every path that changes status must adjust it inside the same transaction, and a periodic reconciliation has to exist, with resource_revision as the authority for what the count should have been.

Follow-up

• A resource moves from archived back to active. Which statements change, and what breaks if the counter update and the status change land in different transactions?
• The cap becomes plan-dependent and a plan can change mid-month. Where does the number 50 live, and who reads it?

Explain the Iron-Carbon Equilibrium Diagram.

Medium · Technical Engineering Fundamentals

Approach

1. Frame it first: this is the metastable Fe-Fe3C diagram, carbon content from 0 to 6.67 wt% (pure cementite) on the x-axis and temperature on the y-axis. Steels sit below about 2.14% C and cast irons between 2.14% and 6.67% C.
2. Name the phases and their structures: alpha ferrite (BCC, at most about 0.022% C at 727 C), austenite (FCC, up to about 2.14% C at 1147 C), delta ferrite (BCC, high temperature) and cementite Fe3C (hard, brittle, 6.67% C).
3. Give the three invariant reactions with their points: peritectic near 1495 C and 0.17% C (delta + liquid to austenite), eutectic at 1147 C and 4.3% C (liquid to ledeburite, austenite + cementite), eutectoid at 727 C and 0.76% C (austenite to pearlite, ferrite + cementite).
4. Walk one alloy through slow cooling with the lever rule. For a 0.4% C steel just below 727 C, pearlite fraction = (0.4 - 0.022) / (0.76 - 0.022), about 0.51, and the rest is proeutectoid ferrite. Name the critical lines A1, A3 and Acm and say that heat treatment temperatures are read off them.
5. State the limit: the diagram assumes equilibrium (slow cooling). Quenched structures such as martensite and bainite do not appear on it; they need TTT or CCT diagrams.

Follow-up

• Why does martensite not appear on this diagram, and which diagram would you use to predict it?
• Using the lever rule, what fractions of proeutectoid cementite and pearlite does a 1.2% C steel have just below 727 C?
• How do hypoeutectoid and hypereutectoid steels differ in microstructure and in mechanical properties?

Describe the Stress-Strain Curve and its significance in material testing.

Medium · Technical Engineering Fundamentals

Approach

1. Describe the test: a tensile specimen is pulled at a controlled rate, with engineering stress = F / A0 and engineering strain = change in length / L0 recorded until fracture.
2. Walk the curve in order: linear elastic region (slope = Young's modulus E, about 200 GPa for steel), proportional and elastic limits, yield point (upper and lower yield in mild steel, or the 0.2% offset yield for metals with no distinct yield), strain hardening up to the ultimate tensile strength, then necking and fracture.
3. Say what each feature gives a designer: yield strength sets the allowable stress (with a factor of safety), UTS is the maximum load capacity, elongation and reduction in area measure ductility, the area under the elastic part is resilience (sigma_y^2 / 2E) and the total area approximates toughness.
4. Explain why the engineering curve falls after UTS: the load drops as the neck forms but stress is still divided by the original area. True stress = sigma(1 + epsilon) and true strain = ln(1 + epsilon) hold only up to the onset of necking.
5. Contrast curve shapes: ductile steel shows a long plastic region, brittle materials such as cast iron fracture with little or no plastic strain, and the curve shifts with temperature and strain rate.

Follow-up

• How do you find the yield strength of a material that has no distinct yield point?
• Why does the engineering stress fall after the ultimate tensile strength while the true stress keeps rising?
• Calculate the modulus of resilience for a steel with 250 MPa yield strength and E = 200 GPa.

How do you calculate mass density in industrial applications?

Medium · Technical Engineering Fundamentals

Approach

1. Start from the definition: density = mass / volume, SI unit kg/m^3 (1 g/cm^3 = 1000 kg/m^3). Separate it from specific gravity (ratio to water, dimensionless) and specific weight (rho times g, in N/m^3).
2. For a regular solid, weigh it and compute volume from measured dimensions. For an irregular solid, use Archimedes: rho = m_air / (m_air - m_submerged) times rho_water. For liquids, name a pycnometer, a hydrometer or an oscillating U-tube density meter. For gases, use the ideal gas law, rho = PM / (RT).
3. Show the industrial use, which runs the other way: mass = density times volume for estimating material and lifting weight. A steel plate of 2 m x 1 m x 10 mm is 0.02 m^3, which at about 7850 kg/m^3 comes to about 157 kg. A pipe or shell is rho times pi(Do^2 - Di^2) / 4 times L.
4. Cover the edge cases: density changes with temperature (and with pressure for gases), porous castings have a bulk density below the true density, and for a mixture with no volume change 1 / rho_mix = sum of w_i / rho_i using mass fractions.
5. As a software engineer, mention how you would code the calculation: carry units explicitly, validate that dimensions are positive, and keep a material density table rather than hard-coded numbers.

Follow-up

• How would you measure the density of a porous casting, and how do bulk density and true density differ?
• How does the density of a gas change when you double the pressure at constant temperature, and when does the ideal gas assumption fail?
• A vessel shell is 3 m long with an outer diameter of 1.2 m and 12 mm walls. Estimate its steel mass.

Can you explain the relationship between material properties and structural integrity?

Medium · Technical Engineering Fundamentals

Approach

1. Organise the answer as failure mode to governing property: yielding to yield strength, excessive deflection or buckling to stiffness (E and the section's moment of inertia), brittle fracture to fracture toughness, repeated loading to fatigue strength, high temperature to creep resistance, and wall loss to corrosion resistance.
2. Give the formulas that make it concrete: Euler buckling load P = pi^2 EI / (KL)^2 depends on E, not on strength. Stress intensity K = Y sigma sqrt(pi a) is compared with K_IC to decide whether a crack of length a grows.
3. Explain the trade-offs: raising strength through carbon content or heat treatment usually lowers ductility and toughness. BCC steels have a ductile-to-brittle transition, checked with Charpy impact tests at the minimum service temperature.
4. Bring in manufacturing effects: welds and heat-affected zones can have different properties from the base metal, residual stresses add to service stress, and post-weld heat treatment and non-destructive testing address both.
5. Close with design margins: a factor of safety on yield or tensile strength, a corrosion allowance added to wall thickness, and inspection intervals that track property loss over the service life.

Follow-up

• Why can a slender column fail well below its material's yield strength?
• Why can a higher-strength steel be the worse choice for a component that sees cyclic loads or low temperatures?
• How would you decide whether a crack found during inspection is acceptable?

What are the primary safety considerations when dealing with static equipment?

Medium · Technical Engineering Fundamentals

Approach

1. Define static equipment (pressure vessels, heat exchangers, columns, storage tanks) and split the answer into design, fabrication and testing, and operation and maintenance, so the answer covers the full life cycle.
2. Design: set design pressure and temperature with margin above operating conditions, design to a recognised code such as ASME Section VIII for pressure vessels, add a corrosion allowance to wall thickness, and check the minimum design metal temperature against brittle fracture.
3. Fabrication and testing: use qualified welding procedures and welders, apply NDT (radiography or ultrasonic) to welds, use post-weld heat treatment where the code requires it, and pressure test before service. Hydrostatic testing is preferred to pneumatic because a compressed gas stores far more energy if the vessel fails.
4. Operation: provide overpressure protection with relief valves or rupture discs sized for the worst relief case, and monitor wall thickness by ultrasonic measurement, using risk-based inspection intervals.
5. Maintenance: isolate, depressurise, drain and purge the vessel, then apply lockout/tagout and positive isolation (blinds). Before confined-space entry, gas-test for oxygen, flammables and toxics, with a permit to work and a standby attendant. Add the controls for hot work, work at height and heavy lifts.

Follow-up

• When would you choose a rupture disc over a relief valve, or use both together?
• Why is a hydrostatic test safer than a pneumatic test at the same pressure?
• Walk through the steps before a person enters a vessel for internal inspection.

Read latency spikes on a sixty-second sawtooth

Medium · cache stampede · single-flight · ttl

The cached listing read path serves about 14k reads/second at an 85% hit rate. p99 sits at 35 ms for 57 seconds, jumps to 900 ms for 3, and repeats. During each spike the primary shows several hundred identical listing queries starting within the same millisecond, all carrying one large tenant's id. Cache entries use a 60-second TTL. Give the mechanism, the ordered checks, the fix, and the correctness hazard your fix must not introduce.

Approach

1. Match the period to a configured number before theorising about load. A spike every 60 seconds against a 60-second TTL is an entry expiring, and you confirm it by correlating spike timestamps with the entry's write time rather than with the traffic curve. If the period had matched a cron or a GC interval instead, this is a different investigation.
2. Establish the concurrency of the miss. Several hundred identical queries in one millisecond means the miss path has no coalescing: every request that arrives between expiry and repopulation recomputes. The herd size is that key's arrival rate times its recompute time, so at 1.2k reads/second for the hot key and a 250 ms recompute you expect about 300 concurrent misses, which matches what is observed.
3. Add single-flight on the miss path so one caller per key recomputes under a short-lived lock while the rest wait for its result. Prefer stale-while-revalidate where the read tolerates it: return the expired value immediately and refresh asynchronously, which removes the latency spike rather than serialising it into a queue of waiters.
4. De-synchronise the keys. Write TTLs with jitter, for example 60 seconds plus or minus 10%, so a deploy or a mass invalidation does not align every key on the same second and turn a per-key herd into a fleet-wide one.

Follow-up

• The same sawtooth appears on a key that is invalidated on write rather than expired. Is that the same bug?
• How does your answer change if the recompute takes 4 seconds instead of 250 ms?

Suggested practice checklist

The plan follows the reported loop: first the screening topics (iron-carbon, stress-strain, density, material behaviour, static equipment safety), then the live-scenario skills, then the commitment questions, ending in a full mixed run.

Day 1 — Map the loop and the topic list

• Write the three reported stages and mark what each requires: referral and application, aptitude test plus technical exam with a 40% cutoff, then the live technical interview
• List the five reported topics (iron-carbon diagram, stress-strain curves, static equipment design, engineering estimation, EHS) and rate yourself red, amber or green on each
• Write a three-sentence description of the job in your own words from the reported duties: estimation tools, electrical maintenance support, safety and quality documentation

Deliverable: A one-page map of the three stages, a red/amber/green topic grid, and your three-sentence job description.

Day 2 — Iron-carbon equilibrium diagram

• Draw the iron-carbon diagram from memory with axes labelled (weight percent carbon against temperature) and compare with a textbook
• Label ferrite, austenite, cementite and pearlite, and mark the eutectoid point (about 0.76 wt% C at 727 °C) and the eutectic point (about 4.3 wt% C at 1147 °C)
• Explain aloud the cooling path of a 0.4 wt% carbon steel from austenite to room temperature, then link carbon content to hardness and ductility

Deliverable: A redrawn labelled diagram and a recorded spoken explanation of one cooling path.

Day 3 — Stress-strain curve and material behaviour

• Sketch the engineering stress-strain curve for mild steel and label the elastic region, yield point, ultimate tensile strength, necking and fracture
• Write what each reading tells a designer: Young's modulus for stiffness, yield strength for the allowable-stress limit, elongation for ductility, area under the curve for toughness
• Prepare a short answer on how material properties relate to structural integrity, including safety factor, brittle versus ductile failure, and what changes at low temperature

Deliverable: A labelled curve and two spoken answers: significance in material testing, and properties versus structural integrity.

Day 4 — Density and engineering estimation

• Write the density formula with SI units, and work three conversions, such as g/cm3 to kg/m3 and the mass of a steel plate from its dimensions
• Work a capacity-style estimate end to end: list assumptions, formula, numbers with units, and a sanity check against a known value
• Do a timed set of quick calculations to build speed for the aptitude portion

Deliverable: Three worked density problems and one estimation walkthrough with assumptions and a sanity check.

Day 5 — Static equipment safety and failure analysis

• Write a checklist of safety considerations for equipment that has no moving parts, such as vessels and tanks: pressure and relief, corrosion and inspection, isolation before maintenance, hot work and confined space, protective equipment
• Practise a failure-analysis answer: loading, stress against yield and tensile strength, material condition, probable failure mode
• Say both answers aloud once and cut anything you cannot explain

Deliverable: A safety checklist and a recorded failure-analysis walkthrough.

Day 6 — Commitment and motivation answers

• Decide your answer on the 3-year bond and write the clarifying questions you would ask about repayment, start point and written terms
• Prepare one pressure story with a check you kept despite the deadline, and the person you told when the plan slipped
• Write your why-Anup answer from the reported work, then a list of the employment terms you would clarify before accepting

Deliverable: Four rehearsed spoken answers: the bond, high pressure, why Anup Engineering and clarifying terms.

Day 7 — Mixed dry run

• Run a mixed set aloud: two technical fundamentals questions and two commitment questions, with no notes
• Repeat any answer where you lost a unit, skipped a consequence or sounded generic
• Write two questions to ask about the team, such as what the software team builds for estimation or maintenance

Deliverable: A completed mixed run, a list of answers to repeat, and two questions to ask.

Commitment, pressure and fit answers

The reported commitment questions cover long-term fit and conduct in an industrial setting: whether you will accept a 3-year bond, how you work under pressure in manufacturing, why you picked this company, and how you handle employment terms. Prepare honest answers with specifics, not polished slogans.

1. Tell me about a time you did not know a technical answer and how you worked toward one.
2. Describe a time you kept a safety or quality check in place when time was short.

Are you willing to work with a 3-year bond?

Medium · Behavioral & Commitment

Approach

1. Decide your real answer before you walk in: yes, yes with conditions, or not without more detail. A bond is a binding commitment, and an answer you cannot stand behind will surface later in the offer stage.
2. If your answer is yes, say why in terms of your goals: you want to build depth in an industrial domain over several years, which a multi-year commitment supports. Keep it to your own reasons and make no claims about the company.
3. If you need information first, say so directly and name what you would ask: what the bond covers, what happens if you leave early (repayment amount, notice, training cost), whether it starts at joining or confirmation, and where the terms will be written.
4. Do not dodge, do not agree on reflex, and do not argue against the bond. A short, honest answer plus one specific clarifying question reads as commitment and as care.
5. Back the answer with a past example of a longer commitment you kept, such as a multi-year project or a degree, to show that you follow through.

Follow-up

• What circumstances would make you want to leave before the three years end?
• What would you need to see in writing before you accept?
• Where do you see yourself at the end of the bond period?

How do you handle high-pressure environments in a manufacturing setting?

Medium · Behavioral & Commitment

Approach

1. Use a real story, ideally from a deadline, outage, lab, plant visit, or a project crunch. If you have no manufacturing experience, say so in one clause and use the closest software or lab example rather than stretching it.
2. Structure it as situation in two sentences, the priority call you made, the actions you took, and the result. Spend most of the time on how you decided what came first and what you refused to skip.
3. Show that pressure did not push you into cutting safety or verification. In an industrial setting, the strongest version names a check you kept (a review, a test, a sign-off) even though time was short.
4. Show communication: whom you told about the slipping date or the risk, and when. Staying quiet while you absorb pressure alone is the failure that sinks this story.
5. Close with a specific result and one thing you now do differently, such as earlier escalation or a standard checklist.

Follow-up

• Tell me about a time under pressure when you made a mistake. What did you do?
• How do you prioritise when two urgent tasks conflict?
• Is there anything you would refuse to rush, even with a deadline on you?

Why are you interested in joining Anup Engineering specifically?

Medium · Behavioral & Commitment

Approach

1. Anchor the answer in plain facts that candidates report: the company works in heavy industrial manufacturing, the software role supports teams for static equipment design, electrical maintenance and safety engineering, and part of the job is tooling for estimation and documentation.
2. Link each fact to something you have done or want to do, such as building a tool for engineers, working with data from physical systems, or an academic project in mechanical or electrical topics. One concrete connection beats three generic compliments.
3. Explain why this setting over a pure software employer: you want your code to affect physical equipment and maintenance outcomes, and you are prepared to learn the engineering behind it.
4. Avoid praise you cannot back up (reputation, size, culture) and avoid reasons that fit any employer, such as location or stability alone. If location or stability matters to you, say it honestly but after the role-specific reason.
5. End with what you plan to learn in the first months, for example the materials and safety basics behind the equipment you would support.

Follow-up

• What do you understand the software team here actually works on?
• Which parts of your background connect to mechanical or electrical engineering?
• What else are you considering, and why does this one rank where it does?

How do you approach clarifying terms and conditions regarding your employment?

Medium · Behavioral & Commitment

Approach

1. Frame it as a routine step of an informed decision. Say you read the offer fully, list anything unclear, and ask in writing before you accept.
2. Name the items you would ask about for this process: the bond (duration, what triggers repayment, any amount), notice period, probation or confirmation terms, work location and site expectations, and how pay is structured. Do not claim any answer in advance.
3. Say who and when: you ask HR once an offer exists or when a term is first mentioned, politely, in an email so that the reply is on record, and you ask for the bond terms before you commit rather than after.
4. Support it with a short example of a time you clarified ambiguous scope or requirements before starting work and avoided a rework or a dispute, since the question rewards a habit, not a one-off.
5. Stay neutral on tone. Asking about terms is not negotiation or reluctance, and you should say what you will do if the answer is not what you hoped: weigh it and decide, rather than assume.

Follow-up

• What would be a dealbreaker in the terms?
• What do you do if the answer you get is vague?
• Describe a time you pushed back on unclear requirements before starting.