◆ Mathematics

What an operations research analyst
really does.

23 tasks, each one witnessed by the sources that watched the job — and behind every one, a prompt you can use tonight.

23evidenced tasks
108,510in the US (2025)
$88,940median pay / year
9systems it runs on
This is what one task looks like here
Present results to management
Present the model results and five-slide executive summary to the fina…3 sources agree

The shape of the day

tap a movement to see its tasks

Which one is you, right now?

Pick the moment · no score, no sign-up
Which moment is you right now?
Whichever you pick, the task behind it opens below.

The work, task by task

23 tasks
Hands on the work15
Present results to management+
Present the model results and five-slide executive summary to the finance director and deputy COO on Thursday afternoon, emphasising decision tradeoffs, budget impacts under current tax rules, and one recommended action with risks and next steps.
jdonetwiki3 agree
when the reply comes backPush once: ask it to sharpen the weakest part, and to say what it assumed. Helpful?
Collaborate with organization members+
Run a two-hour workshop with the procurement lead, tax counsel, and city program manager next Tuesday to walk through assumptions, collect missing constraints, and assign who will validate demand and budget numbers before final analysis.
jdonetwiki3 agree
when the reply comes backPush once: ask it to sharpen the weakest part, and to say what it assumed. Helpful?
Analyze information about alternative actions+
Produce a comparative table of three policy options showing cost, service level, legal constraints, and distributional effects, flagging assumptions tied to current tax legislation and which numbers are sensitive to political changes.
jdonetwiki3 agree
when the reply comes backPush once: ask it to sharpen the weakest part, and to say what it assumed. Helpful?
Create simulations and predictive models+
Build and document a stochastic simulation that projects demand and cost under seasonal variation and policy shocks, include parameter sources, confidence intervals, and a short section on model limitations for the audit file.
jdonet2 agree
when the reply comes backPush once: ask it to sharpen the weakest part, and to say what it assumed. Helpful?
Break systems into components and analyze+
Decompose the service delivery system into modules—intake, processing, dispatch, and billing—map flows, quantify processing times and bottlenecks, and recommend two redesigns that cut average lead time by at least 20 percent.
jdonet2 agree
when the reply comes backPush once: ask it to sharpen the weakest part, and to say what it assumed. Helpful?
Optimize resource allocation and logistics+
Run an optimization that reallocates vehicles and staff across districts to meet demand at minimum operating cost while respecting union shift rules and the council funding cap, and prepare a one-page deployment plan for Monday's ops meeting.
jdwiki2 agree
when the reply comes backPush once: ask it to sharpen the weakest part, and to say what it assumed. Helpful?
Define data requirements and validate information+
Define the exact data fields, sources, retention windows and access controls I need for the municipal revenue model, then validate availability and quality against the finance data warehouse and the tax office extracts before Friday so procurement can buy storage if gaps exist.
jdonet2 agree
when the reply comes backPush once: ask it to sharpen the weakest part, and to say what it assumed. Helpful?
Analyze data to identify trends and patterns+
Produce a trend analysis of five years of tax receipts and grant funding to highlight seasonal, policy change, and outlier patterns, flagging which shifts likely reflect legislation, then prepare charts and a one-page brief for the director next Wednesday.
jdwiki2 agree
when the reply comes backPush once: ask it to sharpen the weakest part, and to say what it assumed. Helpful?
Grow the practice5
Watch and assess2
Keep the record1

What the work runs on

named inside the evidenced tasks
7 tasksIBM SPSS Statisticsanalyzes data distributions and generates statistical comparisons across policy scenarios for informed choice
5 tasksGoogle Docscreates shareable slide notes and a one-page executive summary for management review
5 tasksC++implements performant simulations and custom stochastic processes for large Monte Carlo runs
2 tasksLinuxrun scripts to harvest and preprocess academic papers and manage command-line tools for text extraction and citation handling
1 taskApache Hadoopprocesses large historical routing and demand datasets to feed optimization routines for allocation decisions
1 taskAmazon Redshiftquery and compare large finance tables and extract schemas to confirm field availability and quality
1 taskGitHubmanage code, versioning and team review for the prototype and API specification
1 taskESRI ArcGIScreates spatial networks, performs route optimisation and produces maps and tables needed for evaluation

The same task, four heights

this page is height one
ExecuteDo today's task, with fewer mistakesyou are here → ImproveMake it easy for the next person to acceptin the atlas → DecideWork out the right move when it is unclearin the atlas → BecomeLearn the pattern so it stops coming backin the atlas →

Can AI actually do this job?

the honest answer

It can

where it genuinely helps
  • Explain the theory behind the work
  • Draft, tidy and structure your writing
  • Rehearse a hard conversation before you have it
  • Build a study plan that fits your gaps

It cannot

where it stops, completely
  • Be in the room where an operations research analyst actually works
  • Carry the responsibility when the call is wrong — that weight stays yours
  • Notice what no one wrote down: the hesitation, the thing left unsaid
  • Live with the outcome

What the work pays

two countries, two different measures

United States

this exact occupation · BLS 2025
  • $88,940 a year — the middle: half earn more, half earn less
  • The lowest tenth earn near $57,060; the top tenth near $159,910
  • 108,510 people employed in this occupation

Where the evidence lives

open any of it yourself

Close to this work

12 nearby
MathematicsSimulation Engineer25 evidenced tasks MathematicsGame Theorist23 evidenced tasks MathematicsMath Content Writer22 evidenced tasks MathematicsMathematician20 evidenced tasks MathematicsApplied Mathematician20 evidenced tasks MathematicsTheoretical Mathematician20 evidenced tasks MathematicsQuantitative Analyst Quant20 evidenced tasks MathematicsAlgorithm Engineer20 evidenced tasks

Questions people actually ask

You spend most of your day solving concrete problems: defining the question, collecting and validating data, building a model, and presenting results. Expect blocks of time for coding models in C++ or Python, running simulations on Linux servers, and checking outputs in Amazon Redshift or Apache Hadoop clusters.

You also meet with managers and other teams to explain trade-offs, update decision-support tools (sometimes in Google Docs or GitHub), and write short management reports. Some days are more fieldwork — mapping with ESRI ArcGIS — or literature review for new methods.

Start with the basics used every day: GitHub for version control, Linux for running models, and C++ or Python for coding algorithms. Learn SQL for Amazon Redshift and Hive to query large datasets stored in Hadoop.

Next, pick a statistics tool like IBM SPSS Statistics or Python libraries for analysis and learn ESRI ArcGIS if you will work on spatial problems. If your team uses simulations, practice building them on local Linux or cloud clusters.

The U.S. Bureau of Labor Statistics (BLS) reports 108,510 employed operations research analysts. The median annual wage is $88,940. Expect entry roles nearer the lowest tenth at $57,060, and senior or specialized roles up to the top tenth at $159,910, per BLS 2025 data.

Pay varies by industry (defense, tech, finance), location, and tools you know (big-data tools like Hadoop, Redshift, or advanced simulation skills can push pay higher).

Operations research analysts focus on decision-making: building optimization models, time-and-cost networks, and simulations to choose actions that minimize cost or risk. You will work on planning, logistics, and operational strategy rather than primarily building consumer-facing prediction products.

Data scientists emphasize predictive models and production ML; statisticians prioritize inference and hypothesis testing. There’s overlap: you may use the same tools (Hadoop, SPSS, C++), but your end task is recommending optimal actions and preparing management reports.

Yes, AI can speed simple tasks: generate boilerplate code, suggest model structures, or summarize literature. But always verify outputs—AI can invent wrong equations, misuse statistical tests, or miss constraints. Treat AI like a helpful junior: check code on Linux, validate results against known cases, and review assumptions.

Keep sensitive data out of public prompts. Use internal models or on-premise tools when working with private datasets in Redshift or Hadoop, and follow your organization’s data-security rules before sharing anything.

Study linear algebra, probability, optimization (linear and integer programming), and simulation techniques. Learn to code in C++ or Python, plus SQL for Redshift/Hive and basic Linux command-line skills. Courses in supply chain, project scheduling (time-and-cost networks), and GIS are useful.

Practice by building small projects: an inventory optimization using linear programming, a Monte Carlo simulation, or a route-optimization prototype. Put your code on GitHub and make short one-page reports that show how your model informs a decision.

Ability to translate a real-world problem into a precise model — that is, defining variables, constraints, objective functions, and required data. This skill connects the messy real situation (logs, schedules, budgets) to solvable math and code (C++ programs, optimization solvers, or simulation on Hadoop).

If you can consistently specify what data you need, validate it, and write the model that finds the extreme (min cost, max yield), then learning tools like Redshift, ArcGIS, or SPSS becomes much easier.