Change control impact assessment, explained.

Why a second bag supplier can touch thirty-one things in a drug plant

Five steps of a bag supplier change, from the form to the effectiveness check

1. The owner raises the changeOmar, an engineer, asks to qualify a second bag supplier and lists three things it touches.
2. The team assesses the impactDocuments, systems, equipment, specifications, products, filings, stability, suppliers and training: 31 items.
3. Quality approves, usually at a boardReviewers vote, conditions are written down and the QA chair signs before anything starts.
4. The plant makes the changeFourteen tasks in the example. Eleven of them must finish before the first batch.
5. Someone checks that it workedAn evaluation after go-live confirms the change did what it should and harmed nothing.

Step 2 is the subject of this guide. The company, people and figures in this example are fictional.

What newcomers to GMP change control get wrong: searches, "not affected", inaction and approval

Change control, impact assessment, quality unit and effectiveness check defined

Fourteen terms are defined in the one-page summary.

Nine areas of the plant a GMP change impact assessment has to search

In shortEU GMP Annex 15 gives the list: product quality, the quality system, documents, validation, regulatory status, calibration, maintenance and "any other system". In practice a team works through about nine areas and reads six or more record systems. For each item it writes down a disposition, meaning what must happen to it, and the evidence behind that.

Annex 15, the EU's chapter on qualification and validation, says quality risk management should be used to evaluate planned changes "to determine the potential impact on product quality, pharmaceutical quality systems, documentation, validation, regulatory status, calibration, maintenance and on any other system to avoid unintended consequences".6 PIC/S, the international group of GMP inspectorates, adds two questions many forms leave out: the impact on "other products, processes, systems or sites", and on "other change proposals that may be ongoing at the same time".8

The nine areas, where each one hides and who owns the answer

Three system names recur here. ERP (enterprise resource planning) holds materials and purchasing. MES (manufacturing execution system) runs recipes and electronic batch records. LIMS (laboratory information management system) holds lab specifications and results. Among the teams, MSAT is manufacturing science and technology, and regulatory CMC files changes to the chemistry, manufacturing and controls part of an approval.

Four kinds of finding an impact assessment records

ICH, EU GMP and 21 CFR 211: the change control rules in force in November 2026

In shortUS law, in 21 CFR Part 211 of the Code of Federal Regulations, names the quality unit and written changes. EU GMP asks for prospective evaluation, a risk-based assessment and a check afterward. ICH Q9(R1), Q10 and Q12 explain how to do it well. Since 2021 inspectors have had a PIC/S checklist, and the European Medicines Agency (EMA) said in March 2026 that inspectors in the European Economic Area (EEA) will pilot it between 2026 and 2027. EU Chapter 1 and Annex 15 are being amended; neither amendment is final.

What each rule says about a change, in its own words

Part 211 is written for finished drugs. For active ingredients, FDA reads current good manufacturing practice (CGMP) under section 501(a)(2)(B) of the Federal Food, Drug, and Cosmetic Act, as its 2018 letter on a valsartan process change shows, and ICH Q7 is the guide.24,9 Device makers work under their own quality system rules, which this guide leaves aside.

The quality unit approves each change: what US and EU rules require of the signature

In shortIn the US, the quality control unit must review and approve changes to procedures and specifications. In the EU, responsible persons approve under the quality system, and a Qualified Person confirms the impact was evaluated before releasing a batch. No rule we read requires a board. Many plants use one because it brings the experts together.

The rule that makes change control a legal matter in the United States is short. Procedures, "including any changes", are drafted and approved by the units that use them "and reviewed and approved by the quality control unit".15 A separate section gives that unit the "responsibility for approving or rejecting all procedures or specifications impacting on the identity, strength, quality, and purity of the drug product".7 ICH Q7 says the same for active ingredients.9

European text is framed around the system. Annex 15 says changes "should be authorised and approved by the responsible persons or relevant functional personnel in accordance with the pharmaceutical quality system".6 Annex 16 then asks the Qualified Person, who certifies each batch, to confirm that "the impact of any change to product manufacturing or testing has been evaluated".25

Does a pharmaceutical change control board have to exist?

No text read for this guide requires a board. Annex 15's glossary describes "a formal system by which qualified representatives of appropriate disciplines review proposed or actual changes".6 PIC/S expects "relevant subject matter experts and appropriate internal/external stakeholders" in proposal and approval.8

A board meeting is one way to do that. Most plants write down which changes need it. A common split: QA alone for minor changes and a quorum of QA, regulatory and validation for major ones. MHRA's 2016 deficiency examples include a site with "no definition of which moderate level change controls would require a risk assessment and regulatory affairs review and which would not".26

Changes at contract manufacturers and suppliers

A change often starts outside your walls. FDA's quality agreements guidance tells owners and contract facilities to agree which changes each must report and approve, and it lists "components and/or their suppliers", manufacturing processes, test procedures and major equipment.27 ICH Q12 says the organization responsible for batch release "should be aware of all relevant changes".3 ICH Q7 tells active ingredient makers to notify their customers of changes that can affect quality.9

Major, minor and the formality between them under ICH Q9(R1)

In shortICH Q9(R1), adopted in January 2023, treats formality as a sliding scale set by uncertainty, importance and complexity. It says resource limits cannot justify less formality. It also warns that risk scores can be subjective, and adds a new risk to weigh: losing supply of the medicine.

Most plants sort each change as critical, major or minor. PIC/S describes this as common practice and says a simple low-risk change may need only an impact assessment, while complex or significant ones need formal risk assessments that "more substantially address what might go wrong".8 ICH Q7 allows the same sorting, "depending on the nature and extent of the changes".9

The 2023 revision of ICH Q9 gives the reasoning. Formality "is not a binary concept". Three things decide how much to apply: uncertainty, meaning lack of knowledge about hazards; importance of the decision to product quality; and complexity of the process. One sentence deserves a place in every change procedure: "Resource constraints should not be used to justify the use of lower levels of formality in the quality risk management process."10

Subjectivity: why two reviewers can score one change differently

Q9(R1) says subjectivity "can impact every stage" of risk management, and that it arrives through bias, "when risk questions are inadequately defined, and when tools have poorly designed risk scoring scales". It asks teams to address assumptions and to use "relevant data and sources of knowledge".10 That is an argument for a written evidence trail. A reviewer who has read the three most similar past changes starts from fact. Section 7 shows the arithmetic going wrong.

The risk of not changing, and the risk to supply

Q9(R1) adds a section on product availability, and its annex lists change management as a place to "evaluate the impact of the changes on the availability of the final product".10 PIC/S asks for "potential risks with the current state (until changes are implemented)" and "interim controls" until then.8 For a supplier change forced by an allocation cut, these lines belong at the top of the form, because the change is itself the mitigation.

FDA's 2019 task force on drug shortages said that a "failure to implement such updates and improvements can lead to quality problems that result in drug shortages".28 Section 10 gives its figure.

ICH Q12 inside the quality system: lifecycle documents and where each region stands

In shortICH Q12 lets a company agree which details of its approval are binding, so more changes stay in the quality system. That raises the stakes on change control, because Q12's flexibility depends on it. The US recognizes the tools, with its implementation note still a draft. The EU recognized lifecycle documents only from January 15, 2026, and treats them as optional.

Filing categories and each country's clock are the subject of a separate guide on CMC change impact assessment: in the US, a 30-day supplement (CBE-30) against a prior approval; in the EU, Type IA, IB and II. This guide stops at the point where the quality system asks the question. ICH Q10 frames it: "There should be an assessment to determine whether a change to the regulatory filing is required under regional requirements", and even inside a design space, "all changes should be evaluated by a company's change management system".2

ICH Q12 defines established conditions as "legally binding information considered necessary to assure product quality", where any change "necessitates a submission".3 A product lifecycle management (PLCM) document lists them with their reporting categories. Its practical effect on change control is that the same change can need no filing for one product and a filing for another, depending on what each approval made binding. In the EU, an approval is a marketing authorization, and a filed change to it is a variation.

Why weak change control can cost you Q12's flexibility

The EEA pilot EMA announced in March 2026 looks at this directly. Inspectors review a site's change management, and the page ties the review to planned lifecycle variations. A site with no routine inspection planned gets a standalone inspection of half a day or one day. Applications close December 31, 2026, and inspections may run until March 1, 2027.11

A lifecycle document applies only where it was approved, which is why the assessment reads each approval. We found no MHRA text on lifecycle documents as of November 3, 2026, so ask MHRA how one would be treated for a UK product. Sample facts on this point in Part B are fictional.

The arithmetic of one change: items missed, study coverage, risk scores and a first-batch date

In shortFour small calculations run through most major changes: how much the request missed, whether a supplier's study covers your process, how far a risk score can be trusted, and what date everything supports. Here they are on the fictional bag supplier change.

How much did the request miss?

One change proves nothing about your plant, and this one was picked for its reach. The useful number is your own average. Take 20 closed changes, count what the original form listed and what the closed record ended up with, and divide. We found no public benchmark to compare it with. Also count items added after approval, because those are the expensive ones.

Does the supplier's extractables study cover your process?

Extractables are chemicals a solvent can pull out of plastic in a lab. Leachables are what comes out in real use. A supplier's study covers conditions it chose, so someone must lay those beside your process. In the example, the study ran up to 21 days at 25 °C and 40 °C. The cell culture runs 16 days at 36.5 °C, inside that. The harvest, the liquid collected from the culture, can then sit in the bag for up to 30 days at 2 to 8 °C, colder but longer.

How far can you trust a risk score?

A common tool multiplies severity, occurrence and detectability, each scored 1 to 5. Take the recipe step that rejects a new barcode. Two reviewers, same facts:

Illustrative numbers. Q9(R1) warns about "poorly designed risk scoring scales" and says scores "should be based on an appropriate use of evidence, science and knowledge".10 Write the reason next to each score, and define each scale point in words. A regression test re-runs checks to show that what worked before still works.

What date does everything support?

Add the tasks that must finish first and take the longest path. In the example, approval on October 8, 2026 and eleven required tasks put the first batch on November 20, six weeks later.

The filing clock then limits where that batch may go. A CBE-30 is a US filing, called a supplement, labeled "Changes Being Effected in 30 Days". Product may be distributed no sooner than 30 days after FDA receives it.31 If FDA received it on November 6, a changed lot could ship to the US from December 6, though the batch was released earlier.

How change controls fail, in FDA's and MHRA's own words

In shortTeams change things outside the system, never ask what a change could create, leave the quality impact unmeasured, skip validation, forget the people, miss a filing or never check the result. FDA's warning letters and MHRA's inspection examples describe each one. These letters state FDA's findings at the time they were issued.

The like-for-like trap

A team calls a part "identical" and skips the form. A 2010 article in Pharmaceutical Technology retells a 2003 warning letter about replacement pump pistons that were slightly longer, which it says generated metal particles and led to a recall.37 That is the article's account; the original letter was not read for this guide. The lesson holds: only an evaluation can show that a replacement is like for like.

How often FDA and MHRA cite change control, and why no count captures all of it

In shortFDA publishes the Form 483 citations issued from its system, by year, with no change control code, so any count is a proxy. Our count of four rows worded around changes held at 3 to 4 of every 100 drug citations from fiscal 2019 to 2025, and most of it concerns who may change computer records. MHRA's data shows change control among its citations too. A trade analysis counted 1,577 observations over ten years.

A Form 483 is the list of observations an FDA investigator hands a site at the close of an inspection. FDA publishes spreadsheets of the citations issued from its system, and says they are not a comprehensive listing, because some forms are prepared by hand. In fiscal 2025 (October 2024 to September 2025) the Frequency column on the Drugs tab adds up to 2,837 citations (our sum), and the summary tab shows 713 drug Forms 483. FDA separately reports 1,248 drug quality assurance inspections that year.38,39

Why our Form 483 count can miss change control findings

An investigator can write a change failure under whichever section fits the facts. By our sum, the most cited section in fiscal 2025 was 211.22(d), quality unit procedures not written or not followed, with 243 citations. Next came 211.192, investigations, with 236 across its rows. One row under 211.100(a), absence of written procedures, had 162.38 Any of those can carry a change control story, and a reader cannot tell which from the spreadsheet.

Redica, a trade data firm that classifies the text of the observations, reported 1,577 change control observations in human drug GMP over the ten years to July 2022.40

What MHRA inspectors cite at dosage form sites

What an inspector will ask, from the PIC/S checklist

PI 054-1 applies to GMP inspections of medicine and active ingredient makers, and PIC/S says manufacturers may use it "as a guide to help define their internal change management procedures".8 Its four parts make a fair self-audit:

How a missed impact hurts patients, supply and the next filing

In shortThe worst case is an impurity nobody looked for. More often it is a late batch, a late filing or a finding at inspection, and over time a loss of regulatory flexibility. We found no published cost of a missed impact, so section 18 shows how to measure yours.

What FDA's remedies ask for after a change control finding

FDA's letters show what it wants once it has found a problem. A fix to one change rarely satisfies. The agency asks about the system.

• One 2026 letter asks for "a comprehensive, independent assessment of your change management system", covering "your procedures to ensure changes are justified, reviewed, and approved by your QU", the quality unit.32
• Another letter asks for "a description of the steps that must be taken in your change management system before introduction of new manufacturing equipment or a new product".34
• A third judged a response inadequate because it "did not include a systemic review of your change management program".36

How many changes a drug plant or a biologic carries: the few published counts

In shortPublished counts are rare and each measures something narrower than a plant's real volume. A study of TNF-alpha medicines, which are biologics (drugs made from living cells), found seven regulator-visible changes a year per product. Eighteen global companies counted more than 145,000 post-approval changes needing prior approval, across 156 countries in three years. We found no public figure for days per assessment or share found late.

What hides inside a supplier change: materials, labels, MES recipes and training

In shortA "same material, new supplier" change reaches further than most. The material may differ in ways a certificate does not show, the barcode may break a system rule, the tester may hold a recipe for one film, and everyone trained on the old procedure has to retrain. Each has its own rule to read.

Extractables, leachables and USP chapter 665 for plastic parts

A bioreactor bag touches the product for weeks, and a new film means a new chemical profile. USP, the United States Pharmacopeia, publishes General Chapter 665, which covers plastic components and systems used to make drug products and biologics. USP moved its official date to May 1, 2026, in part to consult on whether to make it an applicable chapter.43

USP's page for the chapter says it "is informational and not compendially applicable unless otherwise specified by regulators and enforcement bodies".44 We read that page's preview and USP's notices, not the full chapter, and as of November 3, 2026 we found no later USP notice or FDA statement that changes this. Vendor pages that call the chapter mandatory go beyond those words, so ask your compendial lead how your regulators treat it.

What FDA's table says about supplier and single-use changes

FDA's June 2021 guidance for biologics shows how fine these distinctions run.45

Our CMC guide covers how to make that argument.

When an ERP, MES or LIMS field hard-codes the old supplier

Plant systems carry business rules that nobody remembers until they fire.

FDA's rule requires "appropriate controls … to assure that changes in master production and control records or other records are instituted only by authorized personnel".18 Annex 11 says any change to a computerized system, "including system configurations", happens only "in a controlled manner in accordance with a defined procedure".46 A useful finding names the step: "MES impacted" gives validation nothing to test.

How much to test is a risk-based judgment, and GAMP 5, ISPE's industry guide, is the usual reference for it.47 FDA's Computer Software Assurance guidance, final on February 3, 2026, is written for medical device production and quality system software, so a drug plant may learn from its thinking without it being a rule for that plant.48

Why a revised procedure means retraining before the first batch

A revised procedure leaves everyone on the old version until they retrain. FDA's 2026 letter on a visual inspector qualification procedure described inspectors continuing critical visual inspections for about three months after the revision without meeting the new standard.35 List who must retrain as part of the impact, and treat it as a condition of the first batch. A separate guide covers procedure revisions and training.

Where AI helps a change impact assessment, and the line draft Annex 22 would draw

In shortAI suits the finding: reading thousands of records for what a change touches and quoting them. People keep the judgment and the signature. We found no regulator text written specifically about AI in change control, and no published accuracy study. Draft EU GMP Annex 22 is the rule to watch, since as drafted it would keep generative AI out of critical GMP uses.

Data integrity for AI-found impacts: ALCOA+, Part 11 and citations

An AI-built list of impacts fed into a GMP decision is a record. MHRA's guidance uses ALCOA, for attributable, legible, contemporaneous, original and accurate, and says the "+" adds complete, consistent, enduring and available with "no difference in expectations".56 21 CFR Part 11, FDA's rule on electronic records and signatures, asks for secure audit trails and signatures that show their meaning, in sections 11.10(e) and 11.50.57 For AI, that means three things on each finding:

These are duties on the company using the system. No product can be certified for them.

Four limits to test before a change board trusts an AI list

• A past change shows what happened then. A similar change that needed no filing in 2021 may need one under 2026 rules. Each suggestion should still point to current text.
• A confidence score is the system's own estimate. Test it against changes where you know the answer.
• Missed items stay invisible. An AI list looks complete, and people may stop looking. Measure recall, the share of real impacts it found, by hiding known items in past changes.
• Records outside the index do not exist for it. Local dossiers on a laptop will not appear. Say what was searched on every assessment.

How QA, validation, QC and regulatory assess a GMP change today

In shortA change owner fills in a form in the quality system. QA routes it to four to six reviewers, each with their own systems. By our estimate, most of the elapsed time is waiting for people and searching for documents; we found no public measure. The ways of doing it vary from paper and shared drives to a module in a quality management system (QMS), with consultants brought in for large programs.

Who takes part in a pharmaceutical change control

Paper, a QMS module or consultants: the usual set-ups for change control

CC-26-1184 from a three-item request to a signed plan with a first batch on November 20

In shortThis is the example from the top of the page, inside the product. Omar's request lists three items. Dana, the QA lead, presses one button, and agents, AI helpers that each do one job, grow it into an impact tree, a branching map of everything the change touches, with 31 cited items. Two of them come only from past changes. Dana accepts the high-confidence items, named people take the three low ones, the board votes, she signs, and the plan follows.

1

The request lists three items

Omar Haddad, an MSAT engineer, raised CC-26-1184 on October 2: qualify Tessmer Single-Use as a second 2,000 L bioreactor bag supplier for Veltrimab and Dravenzo drug substance (the active ingredient), because Wrenmoor Bioprocess cut its 2027 allocation by 40%. It is a Major change, set for Thursday's board. Section 4 of his request names SOP-USP-031, SOP-USP-044 and a US CBE-30 for Veltrimab.

The screen shows exactly that: the change at the center and three dark leaves. The right side already lists the three most similar closed changes.

2

Agents search in parallel

Dana presses "Grow the impact tree". The Change Reader turns the request into what changes, where and for which products. Four agents then search at once: 2,412 procedures and forms, 318 validated systems and 604 equipment items, 1,904 specifications, and 412 registration dossiers across 38 markets. Branches appear as they work.

3

Thirty-one items, in nine areas

When the sample run ends, the tree holds 8 documents, 3 systems, 3 equipment items, 1 specification, 2 products, 10 market filings, 1 stability commitment, 2 supplier items and 1 training curriculum for 56 people. Of these, 26 were found in the sources, 3 came from the owner and 2 from past changes. Three more were checked and cleared.

4

A filing no document names: Japan

The Past-Change Comparer ranks 6,830 closed changes and reads the top three. CC-24-0611, a second sterilizing filter supplier for Veltrimab and 81% similar, needed a Japan minor change notification: Japan's approval matters name the membrane material, and the request had not listed Japan. Veltrimab's Japan approval reads "polyethylene-based bag", so a search for the supplier's name misses it.

The item sits at 71% confidence, so it is marked "your call" for Lena Ortiz, the regulatory CMC lead.

5

The filing a past change missed: the UK

CC-25-0219, a second tubing set supplier for Dravenzo and 78% similar, needed no filing in the US or EU under the product's lifecycle document. In this sample, the UK approval had not taken that document. The UK Type IB was found four months after implementation, at renewal, and filed late with a deviation. The comparer adds the same branch here, citing the past change. It can only add. It never removes a branch another agent found.

6

Recipe step 4.2 and the barcode it would reject

Every branch opens the passage that proves it. Recipe VLT-DS-USP v12, step 4.2, accepts only one barcode format, so it would reject Tessmer's GS1-format barcode labels. How wide to run the regression test is Validation's decision, so at 78% the item goes to Sam Patel. The most similar change, CC-25-0732 at Cork (94%), found this rule and the incoming-inspection procedure only after approval. Its first batch was nine days late.

7

The study that stops at 21 days

The Tessmer film's study ran up to 21 days at 25 °C and 40 °C. Procedure SOP-QC-207 asks for worst-case conditions, and the harvest can sit in the bag for 30 days at 2 to 8 °C. At 62% the question goes to the queue of Priya Ranvelle, the QC lead, who puts it this way: "Does 21 days at 40 °C tell me anything about 30 days at 2 to 8 °C?"

8

The board votes, with conditions in each reviewer's words

Omar agrees. Lena agrees with a shipping restriction per market. Sam agrees with a risk-based regression test of the bag-lot step and the batch record print. Priya agrees on condition that extractables coverage of the harvest hold is confirmed before the first GMP batch. Quorum for a Major change is QA, Regulatory CMC and Validation.

9

The quality unit signs

Dana signs with her password and the meaning "Approved as change board chair". The three open calls go with the decision as conditions, each to its owner, to settle before the first GMP batch: the recipe to Sam, Japan to Lena, the film data to Priya.

10

Fourteen tasks and a release hold

The planner turns accepted items into 14 tasks with owners and dates, each linked to the branch that created it. Eleven must finish before the first batch: documents, the raw material specification, ERP, MES and LIMS, the equipment addendum, the Tessmer audit and quality agreement, the extractables check and training. The first batch on Tessmer bags can be released on November 20, six weeks after approval.

A Tessmer-bag Veltrimab lot still may not ship to the US until 30 days after FDA receives the CBE-30, or to Brazil before approval.

Filing grid, impact matrix, activity record and dashboard QA teams use every week

In shortThe tree is one view of the same evidence. The matrix suits a reviewer working through a list, the grid suits regulatory CMC, the activity record suits an inspector, and the changes and implementation pages suit whoever runs the board.

Who proposes and who approves a change: seven agents cite, and the quality unit signs

In shortEach agent does one job and states what it may not do. None changes a record in your quality system, a system setting or a registration. People accept the branches, confirm each filing, vote and sign. Everything is on the activity record, along with the design targets to test in a pilot.

Dana accepts and signs, Lena confirms filings, the board votes

Items at or above the confidence line, 85% by default, can be accepted together by a person. Items under it go to a named person and are never accepted in bulk. At signing, open calls become conditions. These are controls that support your compliance with your own quality system. The product is never "compliant" or "validated" on its own: you qualify it for your intended use. Botminds' credentials are SOC 2 Type II audited and information security managed to ISO 27001.

The five people on the bag supplier change, in their own words

The records the agents read, and the cited tree, decision and plan they return

Connecting the QMS, document, ERP, MES, LIMS and registration systems

The solution reads copies from your quality management system, document library, registration information system, validation inventory, ERP and training system, and tasks go back to the QMS. Agents read; they change records only through a person. It serves several plants and many markets, with roles, separation of the owner from the approver, and an export of the full record for audit.

Design targets and a sample result to test in your own pilot

Held batches, late filings and reviewer days: the price of a late-found impact

In shortWe found no published figure for this cost, so start from your own records, and this section shows how to read them in a week. The public figures frame the stakes: how many citations touch the subject and how much supply depends on it.

Twenty closed change controls will show you what late finds cost

Unraised changes, repeat misses and board hours no budget line shows

The return on change control automation: reviewer days released and fewer held first batches

In shortPut in your own numbers. The calculator values reviewer days released and, separately, what fewer held batches are worth. Every default is an assumption to replace, and the share you redeploy matters most.

Reviewer days, late finds, send-backs: six levers and how to count each

Your break-even on major changes, with your own reviewer cost and held-batch days

A multi-plant company assessing hundreds of major changes clears the first line on time alone. One site with 40 major changes a year probably does not, and should weigh held batches and filing risk or stay with simpler tools.

Can your closed changes, records and procedures support automation? A QA head's scorecard

In shortPick the answer closest to your site today. Eight questions score 0 to 2, and the last is for reflection. The meter adds up your score as you go, and each question says why it matters and what a strong answer sounds like.

Testing a change control software vendor on evidence, recall and who decides

In shortAsk every vendor the same questions, us included. The answers show whether a tool routes forms or reads your plant, what evidence you keep, how you would measure it on your own changes and how it fits draft Annex 22.

Test it on changes you already closed: a ten-week pilot that ends in a recall number

In shortStart on closed changes where you already know the answer, so you can score recall against a known answer. Then shadow live changes at one plant, with the board still working as it does today. Quality owns the pass line, written down before the first run.

Recall, precision and reviewer days: scoring a change control pilot

Major vs minor, Annex 15, 21 CFR 211.100 and effectiveness checks: the change control FAQ

In shortQuestions QA teams and buyers search for, each answered on its own.