 Heli! Welcomi, wrth gwrs, ymlaen i'n swyddi am ymddangos o'r ffordd, Rhyw hana, rwy'n ddysgol ymlaen yw'r ysgolion. Ac rwy'n gweithio'n gweithio i ddweud i'r ffordd ymlaen yw Rhyw Llywodraeth. Ymlaen yw Llywodraeth is gan ystod. Ymlaen ymlaen yw'r opa. Rha. Ychydig yn ymlaen. Ychydig yn ymlaen 4 wych. 4 wych! Bydd gen幹 rhywbeth. Rydw i'n dechrau drwy cymdeithasen. Tster yn cymoh consecutive. Rym ni'n asa-diacewt iaeth rwyf. Rydych yn gweithio'r ddau o'r gynhau ar gyfer y tîm trafodaethau'u gwahanol, ond mae'r ddechrau. Mae'n fyddech chi'n gwybod mae'n ystod yn cyfnod o'r ddau ffordd mae'r ffordd o ddechrau'r ddau ar gyfer y cyfnod. Ond ydi'r ddau sydd yn rhan o'n mynd i ddim ni i gynnig o'r ddau i wneud yn cael ei gwerthu'r gwahau. Oeddwn ni'n gwybod i'n gweithio, os ydych chi'n gweithio'n gweithio, mae'n gweithio'n gweithio arall o'i gweithio. O ddweud o spennetwork.com i openops.com, spennetwork.com i gweithio'r gweithio'r gweithio i ddweud o ddweud o ddweud i gweithio i gweithio i ddweud, i ddweud i'r angen i'r ddweud i chi, iechyd yn gweithio i chi. Rwy'n cael ei gweld ar-erf. You could put in Norfolk County Council, DWP, and see where their money went, and how much they were spending in different categories over different years. OpenOpps.com is our new site, which is a global site that allows you to search for any tender opportunity in the world, that we've managed to gather so far. We've got data from 80 countries, and it's growing. i talks ond y gbydd. Mae sicrhau bod wedi bod yn gofynir ydych chi'n gweithio. Fe hanes bawb nhw'n gyflwyd, oherwydd mae'n gwahanol yn edrych i gwahanol cwriaeth. Fe hwnnw'n gweithio ar gyfer cyfanyddau. Mae'n gofynir ac yn clys yn ddech ethnicoriaeth, mwyn iddyn nhw'n gweithio am y ddech chi'n gweithio i gweithio ar gyfer cyfanyddau sef aid yn gwybod ar gyfer weithio yna, mae'n gweithio am y cyfanyddau phrychiol a'r cyfinyddau ddim yn ymdae du, a rhai'n mynd i mwyn cyfieithiau wych chi'n dechtyniadau o'r fan ymwreit. Rydych chi'n gyffrous y bysbwysig? Rydych chi'n gwybod y bywbbwysig? Wrth gwybod, ddweud, drwy'n barfyn yn gweithio gyda'r ar-fyn, mae angen defnyddio ar ei fod ei bod yn sicr o gweithio ar greu ei bob wneud i gyda'r ffwynt gdweud o'r bydau i gwybwyll iechyd a'r byw. prompted to be viable to earn a living on all of those interesting things so thats how their work, if you want to use particular search tools on Open Ops or if you want to get more alerts than you have to pay for it I wanted to talk a little bit about why purchasing matters and why it has the potential to change lives Felly, ddyfodd yw'r llwyddo i'n meddwl ar gyfer y cyfledd ymlaen, yn ymdeg o'ch ddau llwyddeddau yn cyfledd yw'r llwyddeddau, mae'n mynd i'n meddwl gyda'r ffaith rhai. Rwy'n meddwl yn ysgrif, i ddwy'r pryd. Mae'n meddwl i'n meddwl ar y cyfledd yma. Felly, mae'r cyfrif iawn yn ei helydd. Mae'n ddim ymdeg arlaen i'ch cyfrif iawn yn y cyfrif iawn. Mae'r cyfrif iawn yn y grannol ar hynny'n o'n ffrif iawn, Cymru yw fwyaf yn ymwneud y salaf o'r adeiladau o'r ysgol. Ac yn ymwyneud yn ddigon fel yw'r amlwysau, o'r amlwysau o'r adeiladau i'w ffagor. Yn ymwneud chi'n cael tuwch. Rhaid i chi rydyn ni yn ymwneud yw'r adeiladau, a dyfodd i gyddiadu i'r adeiladau. Yn ymwneud yw'r adeiladau o'r adeiladau o'r adeiladau, a mae yw, rydyn ni, yn gweithio ar gyfer, yn unig, ydy'r hyn i'r cyflwyniad. Yn cael ei wneud amser ac yn ymwneud ar gyfer y gwirio gael o'r gweithio, mae'n cael ei wneud o'r cyflwyniad i'r cyflwniad ohonydd, ac yn ymwneud o'r cyflwniad i'r cyflwniad i'r gael. mae ydych yn i'r ddweud ar gyfer cyflwyno, ond mae'r problemau ar y profiad yma, y gallai o'r cyfreidio, mae'r cyflwyno yn ymddi'r cyffredinol. Felly mae'r cyflwyno yn ymddi'r cyffredinol rydych yn ei wneud o'r ddweud o'r tachys. mae'r cyflwyno yn ymddi'r cyflwyno mae'r cyflwyno yn ymddi'r cyflwyno. Mae'r ddweud o'r cyflwyno yma fel y rai oír fforddau, yr iawn brybwyr. Wyth wrth dwi'n gwylwch ymlaen nhw. Roedd yn bwas i'r rai a ladech yn pethu cyfath hynny, a'r rai o ffordd yn fawr, maen nhw'n b нап yn bwysig o'r bwysig ymlaen yw cyfrifio'r bwysig, wrth i'r bwysig o'r gwylltig enviwyr o'r bwysig o'r brybwyr. Ddiwrnod byn allan, rwy'n methu o'r hollion i greu eithar ble dydych yn gweld â'r ffrach o croed, ond ie mwyn i chi pa enw. Rwy'n meddwl. Dyna pam ymingau cyflawni nhw y tro bai ewg. Yn gyflawni hwn yn dechrau, ond mae'r hwn yn esconidol, Cyflawni hwn yn ei peth, dw i'r meddwl a gwiaithi, i'r 60 iddyn nhw yn 65, a mae'r 60% gyda'r cefnogi a cyflawni hwn yn cael eu cyfrifau. y creativewer that are reasons so it's easy and fortunately it's pretty easy to do. Conversely, it's hard to detect in that the совет of proof, especially in western societies where you have to get your public prosecutor to go to a court for you. You have to have really strong information that there is fraud occurring. .. mossid yw'r awr ei sgodd yn rhywbeth... ..i mod i feddydd eidio sy'n ddod i yw wneud y ffrodd yn Lly yn ymgledd. Fy yddech chi'n glid i'r hyn yn cyfnod... ..yna'r fflyniad a'r lleidiynau o'r hyn sydd yr hyn yn ymgriff... ..o'r bwrdd yn meddwl i'r dyfodol yn ni yw'r sgawr... ..ym yn neud i'ch sayf yn wiren'n gweithio ar y cyfnod y gallan. Felly, ydych ei bod yn fyrdydd ymlaen... Os ydych yn bwysig, mae'n 50 rhannu yn tyfnwyr o'r cyflosifol yn ei ddweud o'r byd. Felly mae'n gweithio ymlaen nhw yw'r hyn o'r ddweud o'r byd? Os ydych yn 50 rhannu, mae'n gweithio'n ddweud o'r byd yn ymryd yn tyfnwyr i'r cyflosifol yn ymlaen nhw. Mae'n gweithio'n ddweud o'r ddweud, fel fydd yn ddigwydd. Yn y gallu bwysig o'r ddweud, mae'n ddigwydd. Yn ymdweud, mae'r ffordd wedi'u cymhwynt... Yn ymdweud. Yn ymdweud... Yn ymdweud. Yn ymdweud? Yn ymdweud. Yn ymdweud. Mae'r ffordd yn cilio'r ffordd yn cymhwynt fel gweld o'r mynd i'r gwnaeth. Mae'r idea... Crenio, llawer, efallai y mfonsi werth ag ymgyrch yn ym Mhontor, nad ydych yn oed o'i gynhyrch gynharu perthyn, felly dyna'r hwn yn 16 miliwn ar y rhaid i nôl gweld. Yr hwn yn 16 miliwn hanes oes, ac mae'r môl â'r lawer o lot ffonsi waith. Mae'r rhai i llurio yn dechrau eich hwn yn lelwg â nhw. Nid vi'n bwyddo'i'n ffonsi i'r rhaid i'r lwg. Mae'r byth yn dod amser y dyfodol er mwyn i'r byth i'r byth yn edrydd ac rwy'n meddwl gyda y dyfodol sydd yma'r marketig. Efallai mae'r byw gwasiaid ymymu'r fath o boel bod y cyfathirio'i gweld yn ddiwedd i'r myfledd, gyda'r boel i'r boel i'r byth yn awyddfa gafoddiol. Mae'r boel yn fwy o peirlo'r byw, felly mae'n ddif llwyddiol o'r defnyddiol, Because, actually, once you get a... a press report in corruption, you can really see a lack of trust appearing in government as a result. So there's a feeling that it has a particularly high cost. And actually, the best way to tackle it is to prevent it. The problem is it isn't simple. I don't know if any of you have heard of Tom Gash who used to work for the Institute for Government, wrote a has written a wonderful book called Criminal, which is about why people do bad things and about how we look at... Quite we tend to look in the wrong way at how to treat criminality. Either that people are that poverty is the cause of criminality or that for example tougher sentencing will reduce criminality. Well he's really, the book shows that it's about opportunism. If we don't think we are going to get caught we are much more likely to do something criminal. But the really interesting thing around fraud and procurement is that the areas, the blurring of what is criminal and what is not is massive. Rwy'n amdaliad y cyfrifiad o'r cyfrifiadau, mae'r cyfrifiadau cyfrifiadau gofnodd unrhyw, na amdaliad yr hynny'n cyfrifiadau, a mae'r cyfrifiadau bod y cerdfyn arno syniad y cerdd, mae'r cyfrifiadau yn newid y cerdd ymlaen i'r cyfrifiadau cyfrifiadau. Dyma, mae'r greu gnofförau, nw'n digwydd all hynny'n greu greu gnofforddau, ond pe bwyd yn gynnwch i'r ddweud ddiogel ar y dyma, bod hynny'n ddigonwch i'r cyfle cyflwyno, ond y ddechrau'r cyfle am gyfer y gyrfa'r cyflwyno i'r cyflwyno'n cyflwyno, a'r cyflwyno'r cyflwyno yn ymdweithio, ond ond y cyflwyno'n cyflwyno'n cyflwyno, ond nid oes ac yn ffrwd. Maen nhw'n ddigonwch a'r cyflwyno'n cyflwyno. Wrth gwrs, mae'n mynd i'n ddigonwch, You could get the police to go and look at his home computer and do all of those sort of things. But the likely outcome is that you get to sack them and probably you wouldn't even be able to affect their pension. So there's something broken in there. We know it, but how do we do it? Our structures, our governance structures and our procurement laws sometimes make this worse. So we set up framework contracts where we limit the supply to five or six suppliers. That's really a mandate to cartilise your market. Because now those suppliers, they know, they're listed, they can sit down and say, OK, we're going to take it in turns to win bids or are we going to inflate our bids in order to maintain a healthy profit? And you're not going to bid not to win on these. Those frameworks that lock people in really can create that problem. And, of course, selective transparency. We don't release enough data about what's going on in our contracts. That makes it much harder to understand what's going, whether there is crime. The lesson from Tom's book is that actually, if you appear to be monitoring, if you have an always-on presence and say, we are going to be looking at this routinely, you will likely dissuade people from conducting fraud because they think the opportunity isn't there or the risk isn't worth it. And you take away that opportunism. I think the problem we have at the moment is that it is too easy to do it. And one of the other areas that opportunism is quite interesting is, where are we manipulating markets? So if I give pro bono work to a large consultancy, and I'm in one of the central government departments, that scene is a good thing, right? So they're doing a whole lot of work for me for free. Great. Government didn't pay. But actually, that in itself can swing the balance of the competitive nature of any tender that follows because they might, they get to shape the tender quite possibly, but they also get to know exactly how to respond to the tender. You know, these guys don't do pro bono work for nothing, unfortunately. You know, pro bono means free, but it really, they're looking for something in return. So we need to be careful about this. I'm going to run through some stuff around how we can analyse data to try and solve these problems. It's not a perfect world yet, but there's lots of things that you can do to start flagging up issues. So I hope you've, many of you have seen this before, the spend network teapot of finance data or the open ops teapot of finance data. What we're saying is that really we should be publishing transparently all of this data but that it should be linked right down to receipts. I know that's going to be a hard one to persuade people to do, but if you think about it, if I knew that a transaction for software came from a specific budget through a specific contract through a specific tender, I would be able to really understand whether that, whether A, the software was delivering value for money, meeting its intended purpose, but more importantly whether the competition for that software was genuine. Because if I can see it here and I only got say two bid responses on another tender for the same software, there were 15 bid responses, you can start to wonder, well what's going on with that tender that really only elicited two bid responses? So we can start to begin to understand what competition should look like. And the important thing is knowing who bought what from whom. We're still struggling in this country, we don't have a formal register of government entities. Anyone who's met me knows that I'm particularly boring on this subject, but I mean really, come on guys, why can we not know what government constitutes? It would make a lot of what we do around understanding who does what in government really important, really important stuff, make it so much easier. So we do, we can do transaction analysis. Now at the basic level that's like okay, have you got any duplicate invoices this month? You know, invoice fraud is a real thing. People send in duplicate invoices in the hope that they're going to get paid twice, or they send false inv, invalid invoices. So people who send in invoices, the old scam was that you would send in invoices for advertising and the person receiving them go, I didn't advertise on anything. And then they get really threatening letters. So I think, oh actually I better just pay it, it's not that much, it's 300 quid. And then they just do it more and more and more. And then say right, we're going to take you to court and scare people. But actually, you know, payments to suppliers that can't be identified, unknown suppliers, missing approvals, all of these things can indicate fraud. Contract and value for money compliance, exceeding contract values, or payments outside of normal value for money expectations. There is a law, the EU law which still applies for the moment, stipulates that if your contract increases by the original estimate by 10%, then you're required to either re-compete or to send an amendment through. And you should have that approved. The interesting thing is that only, really only 30% of the original tenders have contract award notices in the European portals. And of those, only 5% have values attached to them. So if you inflate your value by 10%, who would know? Especially if you didn't publish a contract award notice and that contract award notice didn't have a value on it, who would know, right? So we, and this is where our systems are broken because we create incentives to not report by doing that. And actually, Savaki have got a really interesting law that says, if you don't publish your contract award notice, your contract is invalid. And I think that's probably an approach that more people should take. We can do stuff like supplier profiling. So actually, does the businessman who owns the supplier have any correlation to your HR list? Or business woman, whatever. But I did some work a few right at the start of my career looking at an NHS trust. And we discovered that one of their suppliers seemed a bit expensive. So we talked to the buyer and said, are you sure these are the right products you're buying? He said, yeah, they're the best company I've ever found. They're amazing. So we thought, well, can't understand it. We decided to give the supplier a call. And it turned out to be the buyer's home number. So it wasn't a particularly sophisticated fraud. And when we went round there, his garage was full of the products that he was selling back to himself. So those sort of things can happen. We can test other things. OK, this supplier only has a registered address that's a PO box. Or they have a registered address where it's very high density. There are sort of like 100 businesses registered to this address. OK, or perhaps they don't have any other clients in the public sector. That's a really strong indicator. So if they only trade with one entity, there's a really strong indicator there that actually there's something going on. Now you have to have other indicators as well. And then supply risk profiling. Just stuff like credit risk, geopolitical risk, performance risk. Are they submitting their company house stuff? All of this makes a difference. And when you can build a picture of what the relationship looks like. Whoops, I'm going to go to from here. I knew that was going to happen. There we go. Are your tenders competitive? There's a really, you know, the procurement 101 says if you're going to buy light bulbs say, I want a light bulb with this wattage, this arc of light and for it to last at least 30 months or whatever, street lamps, right? They have this brightness, this arc, whatever. You don't put in your tender, I need Osram bulbs because then you limit your supply options. But of course, if you are intending to manipulate a tender so that you get the best results out of it so that you're achieving your chosen supplier that you have a collaboration with, then actually that's exactly what you're going to do. So can we analyse tender text to try and the specifications to try and see the commonality between tenders that receive fewer bids and tenders that receive more bids? And then competitive responses. So bid responses, the metadata. So, you know, are they coming from the same IP address? Is the collusion going on? You know, a really interesting example that the CMA found was that some people have been submitting tender responses to their tender portal from the, you know, like two minutes apart from the same IP address. I think that's suspicious. And then the viability of the bid is all the similarity of the bid responses. If you're not going to win, you know, would you just throw stuff in the same as last time and just change the names? Tender patterns, who wins what and when, correlation with employee records. So is this supplier following this buyer wherever they go? This training company, oh, he was in planning and we use this training company and now he's in social care and we're using this training company. Is there anything in that? Contract analysis is the stuff around where your transaction is going. Are we overspending? Are they supplying rebates? And are we getting failures? Collusion is, that's the one that really happens. I think is where we've seen, certainly in the NHS, we've seen a couple of examples recently, a training company. And they had this one of the complex web of money going through three companies. But it was a training hospital in, a teaching hospital in the north of England. And the procurement manager was basically funneling money back to himself and the owner of this training company. And they managed to get through something over 100,000 pounds in three years and no one thought to look. But actually they would have seen very clear relationships with the buyer that existed outside of work. They would have had, there were very uncompetitive tenders. So the relationship with the losers, they were having to get new people in every time they'd issued a tender to really say, oh, bad luck, you never won. Even though they had no intention winning. And there was no value for money. Comparatively compared to everyone else in doing training, they were spending three, four times more. Cartels, very hard to detect. But again, that's about suspicious pricing patterns and bidding patterns and documents. I could go on for too long about that. The one thing I will say, and this is the important bit around gathering a wide set of data, is that you need comprehensive and consistent analysis. I can't compare what is the right number of bids for a piece of software, a piece of CRM software if I haven't got a history of seeing those bids coming up in my territory. So we need that long-term consistent data that is transparent and available to be able to say, oh, we're seeing an average of six bids for CRM on everything. Why did you only get one bid? Our data sources, there's so much inconsistency. So the valuations of contracts, no one ever stipulates whether they include VAT or not. Some supply doesn't include VAT, so you could have understand why would it. But actually unless you're explicit in your metadata about what your valuation is, we get a swing of 20% either way. That makes it very hard to understand what value for money looks like because no one, one person is quoting, oh, well it's a 16 million pound contract. That includes VAT and then someone else is quoting a 16 million pound contract and that excludes VAT. So we need that consistency. And data sharing within government, so some of the data you need to do this stuff is like bid responses or HR data. But you need it across multiple sites in order to understand what the norm is. And I perhaps shouldn't have been surprised, but I was, at how unwilling government is to share its data with other bits of government. There are problems around FOI and the fear that once you give your data to a third party, then they're subject to FOI and you know, are you going to get in problem with your, trouble with your suppliers? But it is important. And we don't have enough on data sharing. We also don't have enough on baselines. Whoops, sorry. I'm going to carry on. Just sorry about this. We don't actually have enough on baselines. So we don't know what like for like analysis is. That VAT is an example, but you need to be able to assess, okay, this person and this person are actually comparative. You know, we can compare what they're doing. And that comes back to identify as other things like that. But the other thing is that our scoring needs to be cumulative. So if someone's doing one thing wrong, you know, let's say they, okay, they only trade with one, they only trade with one public sector entity. That's not enough. We need to know that they've got a PO box, that they are less value for money, that there was only one bid in the tender when they won it, or that they're just sitting suspiciously underneath the threshold. Those are the kind of things that need to be analysed to be able to understand and say, right, okay, that's what you need to look at. And until we do that, we don't get the persuasion element that says you're being watched when you are procuring, and that will put off people's intent on corrupting the process. And most of all, it's, you know, is that give me your metadata? I want your metadata, which is the basic stuff around, actually, what was this? What is this piece of data? Where does it come from? When was it released? What does it represent? Too often, we get stuff that supports to be a tender, but it's not. It's a prior information notice. We get data that proports to be one thing, so spending that actually isn't a transaction. It's a transfer. There's a lot of problems. So that's what we need for ongoing monitoring. We're on our journey of implementing the teapot, and we hope that transparency is really what's going to deliver it. But we've got challenges. One of the challenges is we need to use closed data and open data. We don't expect people's HR data to be opened. That would be wrong. We know that, but it's a useful tool for analysing corruption. So how do we need those restrictions on some pieces of data and openness on others? And we haven't got there yet. And we need business rules that allow us to analyse this data. However, the prize is a great one. So we've been featured in all of these places, so that's my social proof for you. But I would just say that the prize is a great one, and OpenOps is a really exciting opportunity to start understanding what we're doing globally. It is a market of somewhere around between $9 and $11 trillion a year that government spends, and they do that with something around 1.5 million companies. That's actually a scarily small number. If you think about that $9 trillion, it's only going to a million companies. That's a scarily small number, and it has the capacity to change the lives of so many more people, encouraging more smaller companies into the government supply chain, is only going to happen if we have a transparent and genuinely competitive process and well-functioning government markets. So thank you for listening to my spiel. Any questions? Fantastic. I was getting worried. No, we need it for that. Maybe I missed this, but how many data sources do you extract data from in 80 countries? So we're building more data sources every day. I think we've got about, I haven't checked, but I think it's about 600 at the moment. Most of those are from, in terms of volume, the UK has, we have to gather data from over 200 sources in the UK because it's a very fragmented market. Others are a bit better, but we only gather from one data source in the US at the moment, and there are about 1,000. So this is a lifetime's work, really. Working out amongst yourselves. What's the ideal, is the ideal to have all contracts and all that data inputted into one place? Fundamentally, yes. The problem is, so we've gathered what we can in form of tenders and we're going to keep gathering that. If you then gather the contract data, you can then turn around and go, well, what's missing here? So in the UK, contracts finder, just on contracts finder, about 40% of the contract award data is missing. Now, when someone gets round to, when we get round to going, well, who's the one that don't award the contract, don't put the contract awards up the most, we'd be able to rank it and say, well, these are the naughty, you're in the naughty corner. But actually that's really important because then we've, once we got that, you can do things like build pipelines and say these tenders are coming up in the next, you know, 15 months and we can encourage that. That creates more competition in and of itself. So those are, so, yeah, just to collect all the data is the start, but then to analyse it and really to share it with other people. So we're talking to the University of Sussex and to the University of Salford, both of whom, oh and Portsmouth, all of whom have got corruption and counter fraud programmes. And it's like, we want to make that data available to researchers so they can do the best they can with it. Oh, hi, hi. Sorry, a couple of disparate questions. I worked for the EU in the 1980s, so I'm used to the official journal and actually I was one of the people that put the official journal into a large database which kind of still exists. So I'm kind of wondering about data formats is that one of the last things I did because I tend to be a bit of a gadfly is I took the hold of Ter Hamlets spending data for a year which annoyed them intensely and incidentally that was via FOI. When I finally got it, I found that I had CSV, I had XML, I didn't have any JSON at that time but basically I had a whole bunch of disparate stuff which is something that actually also I think a Knowledge Foundation encountered. So I'll actually leave this question there because I'm taking up too much time. So I'm wondering about how you're dealing with that because you've got a lot of different things potentially in there. So we have a mechanism that we go and scrape the location of the files and then we scrape the files down to our servers and then we have to manually check every file because a lot of them have blind headers and the worst of all totals. Total is a real problem because it's also a French petroleum company. So you can't say, okay, we're going to eliminate anything that says total, right? But we want to eliminate something called total. So we have those problems but basically the different data formats we just try and have to cope with it. It involves manual work. There's no way around it. So when we're doing, we've done what? 11,000 CSV files now. We do have quite a clever algorithm so I say it myself that matches the headers. When we do find the headers, it matches the headers using a range of fuzzy matching and machine learning tools to say, oh, that looks like a header that I've seen before and therefore that goes into this column because we previously mapped this into that column. Yes, so actually since all these governments are anxious to show that they're white as snow, there could be actually huge standardisation effort there not that that'll ever happen. Do you know what, I think pigs will fly first. Yeah, exactly. But I would rather not... So one of the things that I see happening is that people stop publishing data because they go, oh, it's got to be perfect. And we would much rather they wrote it on the side of a can, left it in a field where we could. It's better that than not being open at all. So I tend to soft-pedal on data formats. It would be nice if people could do it, but I think it's unrealistic. But also it prevents publication very often. So it'd be better for us to... The one thing I have learnt is that if there's no incentive within for publishing the data, so contract awards is a really good example, there's no real incentive for publishing that data. We have yet to come up with a mechanism for making people do it. And we're supposed to be the best at this. And we're still not able to do it. So I think something around ICO, F-O-Y, being accelerated for these sort of data sets that are pre-approved would be useful. Hi, Ian. Hi, Mark. I'm wondering what the feedback is sort of from the marketplace to open ops and whether you've identified certain categories of potential kind of client. And one I can think of is, I imagine a media-type organisations might be interested in, say, procurement that goes to companies that actually are registered in tax havens and offshore trusts because it would be useful to understand just how much public money is spent on companies that don't pay tax properly in the countries where they receive those commissions. So they might be a certain kind of client for you for specific pieces of work. But what is the emerging kind of profile of the kind of user base for open ops? So for open ops, there's all the people who need to find out about tenders. So that 1.5 million companies would be nice if we had that many clients. But I fear that's not going to happen. But that sort of... I just need to find out what's going on. So that's the core user base. And then we've got both government and I would say organisations that are interested in understanding the competitive nature of public markets. So that might be policy institutions or it might be networks like a media. We're certainly talking to them. I think that it's sort of early days yet. We only launched four weeks ago, five weeks ago. So we're having conversations around that. It's interesting to understand I think how effective procurement markets are and we're feeling aware around that. Government's also interested because they need to understand, well actually, why does this contract go to a small business but this contract not go to a small business? And it has to be more than... Okay, well that contract was worth 160 million and that was worth 16 grand. Okay, we know all of that. But why have I got three contracts for 16 grand and some of them not going to small businesses? What's happening? I was just interested if you'd give an example of international fraud and how you think Open Ops would help that. Like not on a UK level but say in the development context. So I think the idea behind Open Ops is to make the data available. That's the primary. So to give it to people so that they can research it. But let's take the European data and one of the problems we have is varying data quality and inconsistency. But by making the data available on the number of bids received we could start to anyone using... Once that data is in a database you can pretty quickly start running SQL queries on different categories of contracts and identifying ones that have unexpectedly no numbers of bidders. One bid or those sort of things. Now I know there are other researchers doing this as well but what we're trying to do is we're just trying to provide it all in a standard. So when that data does if we start getting that data from beyond Europe and we start getting it from places like Africa then actually and the Middle East you might start to see patterns emerging of particular tenders. Another area that we're looking at for putting a proposal together for a government is to understand where their contracting is occurring. So if I've got a spread of 60 categories where are they contracting most and where are they contracting least? Because every local authority should be buying fuel but they know that some of them just aren't buying any they're not tendering for fuel at all because it's all kind of black market activity. So by spotting where they're not going to tender and where they're not publishing contract awards we can start to indicate where to look further but it's fairly, you know, we would love more data and then we can do more sophisticated things. Does that answer your question? I'm not sure. It doesn't. I mean I do, I have to be careful here because but I know of a specific example in Africa where tender data, not open ops data but tender data is being used to identify where a particular agency are choosing to move money out illegally. So the fact that the tenders are structured in such a way that only one company can win it is being used. But I can't tell you any more than that, I'm afraid. That makes me sound a bit sort of like a secret squirrel but it's just, I can't say anything. James. Thank you. It's always good to be reminded that we're still not great at this stuff so. Well I don't mean, that wasn't... Well no, we're not. So let's be honest. I picked a fight earlier this week with a fellow civil servant who I think needs taking down. Have you got documented like examples where senior civil servants are in collusion with suppliers that have been effectively taken out of office? So the civil servant has been taken out of office? Yeah, you can just Google it. There are people who have gone to jail for colluding with suppliers. There was a case, I think it was 2002, Kent County Council. So not a civil servant but Kent County Council. The guy, they had a program called the Kent Laser Programme which was a collaboration between loads of authorities and what they were doing was they were buying energy. So they went to this one supplier who was doing the mark to market every day to find the best energy prices and providing energy for I think it was 12 or 13 local authorities. So they were all getting prices by the hour and they were getting the good prices. There was a rebate payment that was going back to Kent for the volume of the business and that was supposed to be divvied up amongst all of the councils and this guy was putting it directly into his own bank account which is how they caught him. But yeah, he was sitting on sort of like one and a half million. But his problem was that having started taking all of the money he couldn't put it back because everyone was going to say, where were all our historical payments? What he should have done was only taken a tiny small amount of it and no one would have known. I'm not trying to. But there are good examples. There was an NHS one with the training. There's lots of examples, just Google it. Can I ask what their relationship is with the Crown Commercial Service or whatever it's called now if it has changed its name? Relationship between us and the Crown Commercial Service. Not at the moment, no. I mean we have been, I was in a meeting with them actually a day before yesterday, very lovely team working on small business and they're good and they're working hard but I think they have a hard road ahead of them. But yeah, I mean there are some people who are doing good work at the Crown Commercial Service. Some I would think they should do more to be transparent but we don't have a commercial relationship with them at the moment. It's just a lot of the vocabulary that you have used. I have heard them use the same vocabulary over the years like purchasing energy, things like that. So there's been a lot of commonality here. That's why it was just a recognition. Yeah, no one wakes up on their 11th birthday and says I want to be a procurement professional. But it happened to me and I'm going to make the best of this and I'll try not to use too many of the terminologies that are very procurement-y. But yes, there is a lot of synergy between our work and what they do. We would love for them to talk to us more. But we've got to win tenders and do things like that. Did anyone else have their hand up? I thought I... Well, should we hand those? Yeah, thank you very much guys. Thank you for asking good questions as well. I was worried that it was going to be a bit Bill Hicks for a moment. But thank you. Thanks a lot Ian. We're back on the 2nd of September. We're taking a little break over the summer for ODI Fridays. Thanks everyone for coming and please join me in thanking Ian.