Quantum computing application builder

Quantum computers do not speed up everything. Your first job is to find problems where they may help and to say honestly where they will not.

• Quantum methods are studied for three families of problem: simulating molecules and materials, searching or optimising among huge numbers of choices, and some cryptography tasks.
• Most office work, databases and web pages gain nothing from quantum hardware. A buyer who hears 'quantum' for these should be told so.
• Ask of any problem: how many choices, how costly is a wrong answer, and what does the best classical method already achieve? Write the three answers before you propose anything.
• Count the choices with a simple formula and write the number down. A route that visits 10 stops once each, with the first stop fixed, has 9 x 8 x 7 x ... x 1 = 362,880 possible orders, which a laptop checks in moments.
• Learn the term 'quantum-inspired': classical software that borrows ideas from quantum methods and runs today on ordinary computers. Test it before you recommend buying quantum time.
• Canadian buyers include research groups, banks, logistics firms and energy companies; write one concrete problem for each.
• End every screen with a one-line verdict: use now, watch and look again on a set date, or not suitable. Give the reason in plain words.

Activity. Write a one-page 'is it quantum?' screen for three real problems and give a verdict with the three answers for each.

Standard to meet. Cite the classical baseline in each verdict and reject at least one problem with reasons.

Quantum technology is a controlled field. The rules decide who you may work with and what you may send abroad, so check them before you share anything.

• Strategic technology can need a federal export permit; check the control list before you send code, designs or papers to another country.
• Research funding and public contracts often require security screening of staff and a data-handling plan.
• Cryptography guidance from the federal cyber agency sets timelines for moving to quantum-safe methods; track it.
• Intellectual property: agree in writing who owns code and results when you work with a university or client.
• Cloud quantum hardware may sit in another country. Before you upload client data, read the contract to see where data is stored, and check the federal and provincial privacy rules.
• Foreign students, staff or visitors may need a check before they see controlled designs; check the provincial and federal rule for your case.
• Date every row of your rules table. Lists and guidance change, so set a review reminder every six months.

Activity. Build the rules table for a 3-person quantum start-up in your province, with a source and date per row.

Standard to meet. Give every row an enforcer, action and date, and guess no rule.

You learn by running real circuits on a simulator and on freely available cloud hardware, and recording what the noise does.

• A qubit holds a blend of 0 and 1 until measured. Measure many times ('shots') and you get a pattern of counts, not one answer.
• Start with two circuits: a coin-flip (one gate) and an entangled pair (two gates). Predict the counts, then run them.
• Compare simulator and real hardware: the hardware shows errors. Record the error rate as wrong counts divided by total shots.
• Use version control, and save the circuit, shots, device name, date and result as one table row per run.
• Turn counts into shares by dividing each count by the shots. With 1,000 shots, a count of 470 is 47%.
• Run the same circuit at 100 shots and at 1,000 shots and compare. Write one sentence on why more shots give a steadier result.
• Change one thing at a time between runs, such as the device or the shots, so you know what caused any change.

Activity. Run both circuits on a simulator and on one real device, 1,000 shots each, and tabulate the results.

Standard to meet. Write predictions before runs, compute the error rate and complete the run table.

Quantum work is paid by the project. Price staff time, cloud hardware time and review, and show a buyer what they get.

• Variable cost per package is the hardware time, extra staff review and reporting that each additional study uses.
• Fixed cost per month is the tools, insurance, workspace, training and a share of the lead's time that you pay whatever you sell.
• Offer a fixed-price study (for example 6 weeks) with a clear deliverable: a verdict, a baseline comparison and a plan.
• Add 15% contingency because hardware queues and noise slow the schedule.
• Margin per package = price minus variable cost. Break-even packages = fixed cost per month divided by margin, rounded up.
• Write three cases, with 3, 5 and 8 packages sold in a month, and show the profit or loss for each.
• Check what free or low-cost simulator and cloud access can do before you charge for paid hardware time.

Activity. Build the cost sheet and compute break-even; show a case where only 3 packages sell.

Standard to meet. Show units and currency, cover three cases and check the arithmetic.

Good quantum work is repeatable work. Records make results believable.

• Record device, software version, calibration date, shots, random seed and result for every run.
• Re-run any surprising result three times before you report it.
• Keep a record of expected versus observed counts for each circuit and flag any drift over time.
• Control access to cloud accounts and rotate keys; never paste keys into notes.
• Back up the run table and circuit files to a second place each week, and test that you can open the copy.
• Name files by date and circuit so a colleague can find a run in under a minute.
• At the end of each week, write one page: runs made, surprises, causes found and one change for next week.

Activity. Keep a 4-week run log and one weekly page.

Standard to meet. Fill all six fields in every run and repeat surprises.

Buyers fund what they understand. Learn to explain quantum value in plain words with numbers.

• Use the answer-first form: result, reason, risk, next step.
• Never promise advantage you have not shown; show a measured comparison with the best classical method.
• Offer a no-regret first step: a two-week screen of the client's problems.
• Partner with a university group or industry cluster for credibility and access to hardware.
• Replace jargon with a picture or a number. Say 'tests many options at once' only if you can show how.
• Prepare one slide that shows the classical result, the quantum result and the difference, with units.
• List the three questions a doubtful buyer will ask and write a one-sentence honest answer to each.

Activity. Give a five-minute explanation to a non-expert and collect three written questions.

Standard to meet. Check that the listener can repeat the verdict and the next step.

Quantum advances create real risks as well as benefits.

• Data stolen today may be decrypted later; help clients inventory what must stay secret for ten years or more.
• Plan the move to quantum-safe encryption: list systems, rank by life of data, test, replace in stages.
• Be honest in marketing: do not claim speed-ups that you have not measured.
• Include researchers and users from many backgrounds when you choose test problems.
• Tell clients which of their keys, certificates and vendors would need change, and who in their team owns each.
• Keep client problem data to the minimum needed, and delete it when the study ends as agreed.
• Say what you do not know. A short list of open questions builds more trust than a confident guess.

Activity. Write a one-page quantum-safe inventory for a fictional clinic.

Standard to meet. Give each system a data life and a migration step.

A six-week pilot with a real client proves your process.

• Agree scope, a deliverable and a stop rule in writing.
• Report weekly: tried, learned, next.
• Close with a verdict, evidence and a recommendation, even if the verdict is 'not yet'.
• Plan 12 months: three milestones, funding sources and a hiring plan.
• Set the success measure before you start, such as a measured gain over the classical baseline with units.
• Hold a review with the client in week 2 and week 6 and write down any change to scope.
• List funding routes to check: public research grants, innovation programmes and client co-funding; check each programme's rules and dates.

Activity. Run a pilot on a sample problem and present the review.

Standard to meet. Review includes a verdict, baseline comparison and a 12-month plan.

L1: Technology aware learner. You can define: name the main terms, people and rules of quantum computing application builder. Focus module: Find problems where quantum may help (Understand, 6 h). Quantum computers do not speed up everything. Your first job is to find problems where they may help and to say honestly where they will not. Key points: Quantum methods are studied for three families of problem: simulating molecules and materials, searching or optimising among huge numbers of choices, and some cryptography tasks. Most office work, databases and web pages gain nothing from quantum hardware. A buyer who hears 'quantum' for these should be told so. Ask of any problem: how many choices, how costly is a wrong answer, and what does the best classical method already achieve? Write the three answers before you propose anything. Count the choices with a simple formula and write the number down. A route that visits 10 stops once each, with the first stop fixed, has 9 x 8 x 7 x ... x 1 = 362,880 possible orders, which a laptop checks in moments. Learn the term 'quantum-inspired': classical software that borrows ideas from quantum methods and runs today on ordinary computers. Test it before you recommend buying quantum time. Canadian buyers include research groups, banks, logistics firms and energy companies; write one concrete problem for each. End every screen with a one-line verdict: use now, watch and look again on a set date, or not suitable. Give the reason in plain words.

Speaker notes. Teach: open with the worked example. Delivery routing for 10 stops, with the first stop fixed, has 9 x 8 x ... x 1 = 362,880 orders, which a laptop checks at once. For 40 stops the count is a 47-digit number, but good classical heuristics already find routes close to the best. Verdict: quantum is not yet worth the cost; recommend a classical solver and a watch list with a review date. Then set the activity: Write a one-page 'is it quantum?' screen for three real problems and give a verdict with the three answers for each. Mark against the standard: Cite the classical baseline in each verdict and reject at least one problem with reasons. Check at this level: Name three problem families where quantum methods are studied. Model answer: Simulating molecules and materials, optimising or searching large option sets, and some cryptography. Level note: Start here if new. Evidence: a one-page glossary and a visit or interview note.

L2: Lab or shop assistant. You can describe: explain how quantum computing application builder works from first contact to result. Focus module: Know the rules, exports and standards in Canada (Apply, 6 h). Quantum technology is a controlled field. The rules decide who you may work with and what you may send abroad, so check them before you share anything. Key points: Strategic technology can need a federal export permit; check the control list before you send code, designs or papers to another country. Research funding and public contracts often require security screening of staff and a data-handling plan. Cryptography guidance from the federal cyber agency sets timelines for moving to quantum-safe methods; track it. Intellectual property: agree in writing who owns code and results when you work with a university or client. Cloud quantum hardware may sit in another country. Before you upload client data, read the contract to see where data is stored, and check the federal and provincial privacy rules. Foreign students, staff or visitors may need a check before they see controlled designs; check the provincial and federal rule for your case. Date every row of your rules table. Lists and guidance change, so set a review reminder every six months.

Speaker notes. Teach: open with the worked example. Row: 'Send circuit design to a partner abroad | federal export control | check list, ask for permit if listed | before first send | permit or written no-permit-needed note on file'. If the table has 12 rows and reviews run every 6 months, you recheck 12 x 2 = 24 rows a year, about 2 a month. Then set the activity: Build the rules table for a 3-person quantum start-up in your province, with a source and date per row. Mark against the standard: Give every row an enforcer, action and date, and guess no rule. Check at this level: Why check export rules before sharing designs? Model answer: Strategic technology can need a federal permit to leave the country. Level note: Evidence: a flow chart and a two-minute spoken explanation.

L3: Technician. You can apply: carry out a routine task in quantum computing application builder to a written standard. Focus module: Build and run a small quantum program (Apply, 10 h). You learn by running real circuits on a simulator and on freely available cloud hardware, and recording what the noise does. Key points: A qubit holds a blend of 0 and 1 until measured. Measure many times ('shots') and you get a pattern of counts, not one answer. Start with two circuits: a coin-flip (one gate) and an entangled pair (two gates). Predict the counts, then run them. Compare simulator and real hardware: the hardware shows errors. Record the error rate as wrong counts divided by total shots. Use version control, and save the circuit, shots, device name, date and result as one table row per run. Turn counts into shares by dividing each count by the shots. With 1,000 shots, a count of 470 is 47%. Run the same circuit at 100 shots and at 1,000 shots and compare. Write one sentence on why more shots give a steadier result. Change one thing at a time between runs, such as the device or the shots, so you know what caused any change.

Speaker notes. Teach: open with the worked example. Entangled pair, 1,000 shots, ideal result: about 500 '00' and 500 '11'. Real device: 470 '00', 455 '11', 75 other. The three counts add to 470 + 455 + 75 = 1,000, so nothing is missing. Error rate = 75 / 1,000 = 7.5%. Then set the activity: Run both circuits on a simulator and on one real device, 1,000 shots each, and tabulate the results. Mark against the standard: Write predictions before runs, compute the error rate and complete the run table. Check at this level: How do you compute the hardware error rate? Model answer: Unexpected counts divided by total shots. Level note: Evidence: a supervised task record that meets the standard once.

L4: Product engineer or builder. You can analyse: diagnose why a quantum computing application builder result missed its target and fix it. Focus module: Cost a proof-of-concept project (Evaluate, 8 h). Quantum work is paid by the project. Price staff time, cloud hardware time and review, and show a buyer what they get. Key points: Variable cost per package is the hardware time, extra staff review and reporting that each additional study uses. Fixed cost per month is the tools, insurance, workspace, training and a share of the lead's time that you pay whatever you sell. Offer a fixed-price study (for example 6 weeks) with a clear deliverable: a verdict, a baseline comparison and a plan. Add 15% contingency because hardware queues and noise slow the schedule. Margin per package = price minus variable cost. Break-even packages = fixed cost per month divided by margin, rounded up. Write three cases, with 3, 5 and 8 packages sold in a month, and show the profit or loss for each. Check what free or low-cost simulator and cloud access can do before you charge for paid hardware time.

Speaker notes. Teach: open with the worked example. Study package price 2,400; variable cost 900; margin 1,500 (about 63%). Fixed cost 7,000 a month needs 7,000 / 1,500 = 4.67, so 5 packages. With 15% contingency on variable cost, 900 x 1.15 = 1,035, margin 1,365, and 7,000 / 1,365 = 5.13, so 6 packages. Three cases at the 1,500 margin: 3 packages give 4,500, a loss of 2,500; 5 give 7,500, a profit of 500; 8 give 12,000, a profit of 5,000. Teaching figures in CAD. Then set the activity: Build the cost sheet and compute break-even; show a case where only 3 packages sell. Mark against the standard: Show units and currency, cover three cases and check the arithmetic. Check at this level (this question is covered in module 6: Explain it and win a first client): What is a no-regret first step for a client? Model answer: A short screen of their problems with a stop point. Level note: Evidence: a fault or gap analysis with a fix and a repeat test. This course targets this level.

L5: Senior engineer or line lead. You can evaluate: judge options for quantum computing application builder against cost, risk and quality. Focus module: Explain it and win a first client (Create, 6 h). Buyers fund what they understand. Learn to explain quantum value in plain words with numbers. Key points: Use the answer-first form: result, reason, risk, next step. Never promise advantage you have not shown; show a measured comparison with the best classical method. Offer a no-regret first step: a two-week screen of the client's problems. Partner with a university group or industry cluster for credibility and access to hardware. Replace jargon with a picture or a number. Say 'tests many options at once' only if you can show how. Prepare one slide that shows the classical result, the quantum result and the difference, with units. List the three questions a doubtful buyer will ask and write a one-sentence honest answer to each.

Speaker notes. Teach: open with the worked example. Pitch: 'We screened six scheduling problems; two may benefit from quantum-inspired methods now, which is 2 of 6, or 33%. We propose a six-week study with a stop point after week two.' The client can say yes to a small step, not a large bet. Then set the activity: Give a five-minute explanation to a non-expert and collect three written questions. Mark against the standard: Check that the listener can repeat the verdict and the next step. Check at this level: What should a pitch avoid? Model answer: Promising advantage you have not measured against the best classical method. Level note: Evidence: an options paper with numbers and a recommendation defended to a panel.

L6: Programme or plant manager. You can create: design and pilot a new offer or process in quantum computing application builder. Focus module: Handle security, ethics and trust (Evaluate, 6 h). Quantum advances create real risks as well as benefits. Key points: Data stolen today may be decrypted later; help clients inventory what must stay secret for ten years or more. Plan the move to quantum-safe encryption: list systems, rank by life of data, test, replace in stages. Be honest in marketing: do not claim speed-ups that you have not measured. Include researchers and users from many backgrounds when you choose test problems. Tell clients which of their keys, certificates and vendors would need change, and who in their team owns each. Keep client problem data to the minimum needed, and delete it when the study ends as agreed. Say what you do not know. A short list of open questions builds more trust than a confident guess.

Speaker notes. Teach: open with the worked example. Inventory row: 'Customer records | secret for 25 years | high risk | pilot quantum-safe key exchange in test system by Q3'. A system whose data must last 25 years and whose migration takes 3 years should start planning now. Then set the activity: Write a one-page quantum-safe inventory for a fictional clinic. Mark against the standard: Give each system a data life and a migration step. Check at this level: Why inventory long-life secrets now? Model answer: Data stolen today may be decrypted later by quantum computers. Level note: Evidence: a pilot with measured results over at least 30 days.

L7: Technology founder or director. You can lead: run a team or business in quantum computing application builder and develop others. Focus module: Run a pilot study and plan growth (Create, 8 h). A six-week pilot with a real client proves your process. Key points: Agree scope, a deliverable and a stop rule in writing. Report weekly: tried, learned, next. Close with a verdict, evidence and a recommendation, even if the verdict is 'not yet'. Plan 12 months: three milestones, funding sources and a hiring plan. Set the success measure before you start, such as a measured gain over the classical baseline with units. Hold a review with the client in week 2 and week 6 and write down any change to scope. List funding routes to check: public research grants, innovation programmes and client co-funding; check each programme's rules and dates.

Speaker notes. Teach: open with the worked example. Stop rule: 'If after week 2 no problem passes the screen, we stop and refund the unused part.' In a 6-week pilot, week 2 is one third of the time, so the client risks little. Outcome: one problem passes; study continues. Then set the activity: Run a pilot on a sample problem and present the review. Mark against the standard: Review includes a verdict, baseline comparison and a 12-month plan. Check at this level: What must a pilot agree before it starts? Model answer: Scope, deliverable and a stop rule. Level note: Evidence: a record of results over a year and of people you have coached to a higher level.